The Dog Its Domestication And Behavior

THE

DOG

Its Domestication and Behavior

other books by the author:

Canine Behavior
Canine Pediatrics

Integrative Development of Brain and Behavior in the Dog
The Behavior of Wolves, Dogs and Related Canids
Understanding Your Cat
Understa?:ding Your Dog

Concepts in Ethology: Animal and Human Behavior
Betiveai Animal and Man, The Key to the Kingdom
The Wild Canids: Their Systematics, Behavioral Ecology and
Evolution

THE

DOG

Its Domestication and Behavior

Michael W. Fox

Director, Institute for the Study of Animal Problems
Humane Society of the United States, Washington, D.C.

Garland STPM Press

Nezu York & London

Copyright © 1978 by Michael W. Fox

All rights reserved. No part of this work covered by the copyright hereon may be
reproduced or used in any form or by any means — graphic, electronic, or
mechanical, including photocopying, recording, taping, or information storage
and retrieval systems — without permission of the publisher.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1

Library of Congress Cataloging in Publication Data

Fox, Michael W 1937-
The dog.

Bibliography: p.

Includes index.

1. Dogs— -Behavior. 2. Canldae— Behavior.

3. Domestication. 4. Mammals— Behavior, J, TiiJe-
SF433.F69 599'.74442'045 76-57852

ISBN 0-8240-9858-7

CarUnd STTM Press.

A division of Garland Publishing, Inc.

Contents

Acknowledgments vii

Introduction i

I. Effects of Domestication in Animals: A Review 3

II. Sodo-Ecology of Wild Canids: Environment

and Behavioral Adaptation 21

III. Behavior and Ecology of an Urban Feral Dog

Pack 41

IV. Vocalizations in Wild Canids and Possible

Effects of Domestication 69

V. Behavior Genetics of F, and F; CoyotexDog

Hybrids 9 '

VI. Effects of Domestication on Prey-Catching
and Killing in Domesticand Wild Canids and

F:; Hybrids m-

VI

VII. Interspedes Interaction Differences in Play

Actions in Canids 133

Vin. Sodalization Patterns in Hand-Reared Wild

and Domesticated Canids 141

DC. Stages and Periods in Development:

Environmental Influences and Domestication 153

X. Behavior, Development, and

Psychopathology of Cardiac Activity in the

Dog 177

XL Condusions: Domestication and Man-Dog

Relationships 235

Appendix I 263

Appendix 11 267

Appendix III 271

References

Index

291

Acknowledgments

The author gratefully acknowledges the collaboration of the follow-
ing undergraduate assistants and colleagues in the completion of
some of the material included in this book: Keith Kretchmer (Chap-
ter I), Ellen Blackman and Dr. Alan Beck (Chapter 3), and James
Cohen (Chapter 4). I am espedally glad for the opportunity to
include some of my Army research data collected in collaboration
with Dr. Jeff Linn (Chapter 10) and to reevaluate this material from
the original report (Linn, 1974) in relation to other data derived
from wild canids and their hybrids in my laboratory. Thanks also to
Dr. R. V. Andrews for making the plasma cortisol estimations in
the wolf cubs which are included in this report.

Much of the material incorporated in this text, some of which
has been published separately as short artides in sdentific jour-
nals, constitutes an integrated series of studies which provides a
valuable multidimensional view of domestication and behavior.

My research was supported in part by NSF grant GB-34172,
PHS grant P 10-ES-00139 awarded through the Center for the Biol-
ogy of Natural Systems, Washington University, and a grant from
the Arctic Institute of North America under contractual arrange-
ment with the Office of Naval Research.

Thanks also to Ms. Wanda Meek for her invaluable secretarial
assistance in preparing this manuscript.

Introduction

Over the past 13 years a series of studies on the development of
brain, behavior, and sodalizafion in various breeds of domesH-
cated dog have been undertaken (Fox, 1965 and 1971). This research
was subsequently refocused on wild (undomeshcated) canids, in-
duding the wolf, coyote, and red fox, in order to help identify
changes in behavior and sodalization due to domestication in
dogs (Fox, 1971b). Differences among various wild spedes, in terms
of communication and sodal organization, correlated with be-
havioral adaptation to a particular niche or set of ecological var-
iables (Fox, 1975).

These observations led to the question of the adaptability of
domestic dogs to “revert" to the wild, to become feral and live
independent of man. The reversible and irreversible effects of
domesticahon might then be demonstrated. A study of feral urban
dogs was undertaken to answer some of these questions. This
study underlined the tragic consequences (to the dog) of irrespon-
sible ownership, and several sodal, ecological, and public health
hazards were demonstrated and publicized.

Earlier studies of behavior development and sodalization of
the dog provided the basis for a number of publications to educate
the general public and thereby improve the relationship bctivecn
man and dog through understanding (Fox, 1972). Following some,
but considerably less, research on the behavior development of

2

the dog: its domestication and behavior

cats and an extensive survey of the literature, similar matenal of an
educational nature was prepared and published for the same pur-
pose (Fox, 1974). . , ■ ,

Research on the effects of early expenences (e.g., soaal isola-
tion, restricted socialization) on later behavior served to establish
some parameters for interpreting and treating abnormal behavior
and emotional reactions in adult dogs and other domesticated ani-
mals (Fox, 1968 and 1972). This, coupled with nationwide consulta-
tions generated via the media (radio and television broadcasts and
syndicated newspaper and magazine articles) and referrals of cases
by veterinarians in private practice, produced data on a wide range
of behavioral anomalies in pets, and often related problems in the
pet-owner relationship.

The phylogenetic origins of the dog, established long before
man's intervention, are still evident today, together with the effects
of 10,000-14,000 years of selective breeding during its period of
domestication. As the various needs and life-styles of people
change, so the relationship between pet and owner and pet and
home environment changes. Such contemporary influences repre-
sent the more dynamic and labile aspects of the process of domesti-
cation, an understanding of which is important for the well-being
of animal and man alike.

Some of the more general aspects of domestication first are
reviewed from an ethological vieivpoint and various concepts and
principles outlined. Given the complexities of the domestication

process and the lack of historical records, an objective study of the
effects of domestication in any given species must be conducted at
four levels; comparative behavioral and sodo-ecological studies of
wild and domesticated canids; genetic studies of hybrids of wild
and domestic spedes; sodalization and effects of early experiences;
the development of sodal relationships in hand-raised wild and
domesticated spedes. The strong emphasis in these studies on the
behavior of ivild canids, which may seem irrelevant or super-
fluous, adually provides the only template for identifying and
analj’zing the possible effects of domestication in the dog.
Additional aspects of behavior development relevant to the com-
plex phenomenon of domestication are also reviewed, and some
unexplored potentials for future improvement of livestock and pets
alike is discussed.

I

Effects of
Domestication in
Animals: A Review

Introduction

4

the dog: its domestication and behavior

Table I.

Under noturol conditions ^

1. Emergence of a new character or character complex making it possible to
occupy previously unoccupied ecological niches.

2. Explosive adaptive radiation into all available ecological niches (carni-
vores, herbivores, etc.).

3. Darwinian selection between species.

4. Adaptive radiation within species.

5. Increasing specialization within the ecological niche.

6. Specialization by minor or nonadaptive (neutral) specialization.

7. Cessation of discernible evolutionary change — highly specialized in an
unchanging environment (e.g., horseshoe crab); extinction if unable to
adapt to the changing environment.

Under domestic conditions

1. A shift in human culture provides a new ecological niche for those
species with favorably adapted characteristics.

2. Explosive domestication of many forms to serve human culture (scaven-
gers, milk producers, draught animals, etc.).

3. Darwinian selection between species.

4. Divergence vdthin domesticated spedes toward different specialized
functions (e.g., among dogs, cattle, etc.).

5. Inaeasing improvement along a straight line for a behavioral function
(increased milk production, speed, etc.).

6. Development of breeds or varieties by nonfunctional (morphological)
diversification and imposed reproductive isolation.

7. Some forms apparently do not respond to further selection; the con-
dition is likely to be transient, or the stock replaced by a line which will
continue to respond.

The factor of time over gencratiorrs is a vital aspect of the
process of domestication. Domestication is an evolutionary process
resulting from changes in the selection pressures on a species or
population created by an artificial environment, with release from
the competition for survival characteristic of a natural habitat (Allee

Effects of Domestication

5

E = Experiences, pre & postnatal

G = Genetic influences (adrenal-pit;

neoteny; pseudospec iat i on
* I nfant i 1 i sin;dependency

Figure i. Schema of interrelated factors in the taming and later
domestication of a wild animal. Socialization and related
environmentallexperiential factors and genetic selection are entailed
in producing a domesticated phenotype.

et at., 1949). The significance of such environmental influences are
elaborated upon in this book. Such conditions produce changes in
the gene structure of a population and differential reproduction of
genotypes over generations which could result in eventual specia-
tion. Hale (1962) compares the general evolutionary sequences in
animals under natural and domestic conditions. (See Table I and
Figure i.)

History of
Domestication

Though there is little agreement over the history of domestication,
scholarly inquiry has produced a number of interesting theories.
Zeuner (1963), in observing the different types of social relation-
ships between animal and man, has proposed that domestication

6

the dog: its domestication and behavior

occurred long before the economic potential of animals in agricul-
ture was discovered. He outlines five stages of the intensity of
domestication, which describe both present relationships between
man and animal and the historical development of domestication.

1. Loose contacts with free breeding.

2. Confinement to human environment with breeding in cap-
tivity.

3. Selective breeding organized by man to obtain certain
characteristics and occasional crossing of wild forms.

4. Economic considerations of man leading to the planned de-
velopment of breeds with certain desirable properties.

5. Wild ancestors persecuted or exterminated.

With this outline of social relationship, Zeuner (1963) de-
veloped a probable order in which species were taken into domes-
tication,

1. Mammals domesticated in the pre-agricultural (Mesolithic)
phase.

a. Scavengers: dog, pig, duck.

b. Social parasitism: reindeer, sheep, goat.

2 . Mammals domesticated in the early agricultural (Neolithic)
phase.

a. Crop robbers; cattle, buffalo, gaur, batteng, yak, pig.

b. Those systematically domesticated: fowl, hyena, ostrich.

c. Pest destroyers: cat, ferret, mongoose.

3. Mammals subsequently domesticated primarily for transport
and labor.

a. Domesticated by agriculturalists in the forest zone:
elephant.

b. Domesticated by secondary nomads: camel.

c. Domesticated by river valley civilizations: ass, onager.

4. Various other mammals,

a. Small rodents: rabbit (Medieval), dormouse (Roman).

b. Experimental domestication: hyena (Egyptian), fox
(Neolithic), gazeUe (Egyptian), ibex (Egyptian).

c. New World species: Hama (American Indian).

d. Pets: Mouse (modem European).

5. Birds, fishes, and insects (not classified).

Effects of Domestication

7

Artificial Speciation
and Adaptive Radiation

Under any condition, natural or domestic, the major factor produc-
ing evolutionary change is isolation of one population of a species
from another. In the domestic situation, a population is intention-
ally isolated by man (whereas isolation under natural conditions is
a result of chance). Such isolation occurs on two levels. Geographic
or reproductive isolation occurs where a topographical barrier
prevents two populations of a species from interbreeding. This
leads to divergent adaptation of each population to its environment
through new phenotypic expression of the common genotype. If
the populations are aUowed to interbreed (gene flow) after a rela-
tively few number of generations, speciation will not occur. Should
such gene flow be inhibited by geographic isolation for a long
period of time (in generations), divergence might ocair to t"® ex-
tent of making the two populations reproductive y “e,

and hence separate species. Isolation on this level is twofold. Be-
havioral isolation is the result of divergence of courtship-mating
patterns between populations. Incompatibility may also result
from divergence in the characteristic anatomies ° P°P“'^-

tions making reproduction mechamcaUy

tic strains are reproductively compatible with the wild counte^arts
of their spedes. There are very few speaes of animals that are

solely domestic. ...

It is obvious therefore, that domestication is, as is any evolu-
It IS , ' f intricate combination of environ-

honary process, the 'result ofj^n inmc potentials of a

mental influences coup ® spedfic environment, man

spedes. Through manipulation of the spea ,Dedes to

can accelerate the evolution of certain ° hich

suit his spedfic needs Other enwrOTmen^a reactivity,

may greatly modify the animal s be

and sodabmty, wUI be discussed m Chapter 9.

8

the dog: its domestication and behavior

The Effects of
Domestication on
Behavior

MAJOR INFLUENCES

With an understanding of what domestication involves, we can
begin to look at its various effects. There are a few major factors
inherent in the process of domestication itself that directly influ-
ence the behavior of a species (Fox, 1968c). The first of these is the
change in environments from a natural habitat to the habitat
created by man. The basic evolutionary aspects of this change al-
ready have been discussed. An animal will respond to the type of
domestic environment in which it is placed. An artificial environ-
ment, depending on the specific circumstances, can have a great
influence on the general behavior, disease resistance, and total
productivity of the species. The average zoo, where conditions of
captivity have many of these effects, is a good example of this.
Different individuals and different species will adapt better to this
type of situation than others. A simulated natural environment
frequently is easier to adapt to and provides for a minimum of
incidence of behavioral abnormalities and stress diseases. Man will
create either of these domestic environments to suit his needs (Fox
and Walls, 1973).

The second major factor influencing behavior (Fox, 1968b) is
the genetic selection of specific strains from a few species of ani-
mals originally domesticated. Man selects such strains for a
number of general characteristics.

1. DodUty — this is a heritable trait.

2. Adaplability and fitness for different domestic environments.

3. Desirable characlerislics of economic importance which can be
enhanced through breeding such as high fertility, rapid
growth, efficient food conversion, etc.

4. Fixed paedormorphic features which reduce the time span
from birth to maturity, but allow a high degree of adapt-

Effects of Domestication

9

ability because of perpetuation of certain infantile charac-
teristics.

5. Reduction of wild characteristics (especially aggressive and
sexually related display structures) such as horns, hair, etc.,
by neoteny (similar to 4 above).

6. Hybrid vigor through crossbreeding.

This type of direct (genetic) selection will precipitate a ar
specialized deviation from the natural behavioral nom an
changes in behavior patterns resulting from sirnply a c
environment. Also, any exaggeration of certain p
teristics through breeding will be accompanied y P
changes in characteristic behavior.

CHARACTERISTICS FAVORING DOMESTICATION

A number of natural behavioral stat^

adaptive in faciUtating transition from a wi o ^ domesti-
Hale's (1962) observations are „.,_ricHcs listed on the

cated species might not show all of the cha ^pc showing a

left sid"e of this fable, but it is colfmn cof Id

majority of the behavioral characteristics i &

adapt to a domestic situation.

Group Structure u„.H.ivr>e social groups with

In a domestic situation, flock- or he j, n,ore managc-

often well-defined dominance hierarchies structure,

able and easily maintained than a teiritona , orcanization re-

A loose social hierarchy or leader-follower to the

duces the number of conflict situations a croup would have
social stress of confinement, situations where

a similar advantage over a terriloria to provide large

large numbers must be housed ■ h the domestication of

quantities of produce. (An exception to * proximity mtol-

lerrilorial species as family pels.) A re also have

orance (social distance) and intraspeafic « gg
had high selection priorities in domes ica

10

THE dog; its domestication and behavior

Table 11.

TavoraUe characteristics

1. Group structure:

a. Large soda! groups (Hock, herd, pack), true leadership.

b. Hierarchical group structure.

c. Males affiliated with female group.

2. Sexual behavior;

a. Promiscuous matings.

b. Males dominant over females.

c. Sexual signals provided by movement or posture.

3. Parent-young interactions:

a. Critical period in development of spedes-bond (imprinting, etc.).

b. Female accepts other young soon after parturition or hatching.

c. Precodal young.

4. Responses to man:

a. Short flight distance with man.

b. Little disturbed by man or sudden changes in environment.

5. Other behavioral characteristics:

a. Omnivorous

b. Adapt to a wide range of environmental conditions.

c. Limited agility

Unfmjorable characteristics

a. Family groupings.

b. Territorial structure.

c. Males in separate groups.

a. Pair-bond matings.

b. Male must establish dominance over or appease female.

c. Sexual signals provided by color markings or morphological structures.

a. Spedes-bond established on basis of spedes characteristics.

b. Young accepted on basis of spedes characteristics (e.g., color patterns).

c. Altridal young.

a. Extreme wariness and long flight distance.

b. Easily disturbed by man or sudden changes in environment.

a. Spedalized dietary habits.

b. Require a spedfic habitat.

c. Extreme agjlity.

Effects of Domestication

11

Sexual Behavior

The most important adaptive trait of a species is its ability to
reproduce under domestic conditions. Promiscuous sexual be-
havior has obvious advantages over the establishment of pair-
bonds in domestic situations where certain traits are being bred for,
and sires are being used. Established male dominance over females
reduces conflict between courtship behavior and aggression, thus
normally facilitating mating behavior. Posture and movement as
sexual response eliciting stimuli are more stable and adaptive and
presumably more important than sexual display structures. The
latter secondary sexual characteristics are less necessary for re-
production in captivity since their reduction (through neoteny)
does not lower reproductive success in captivity (although it would
in the whd). Disruption of the former characteristics may directly
affect the breeding capabilities of the species. Other social and
developmental influences on reproductive behavior are discussed
in Chapter 9. (See also Enders, 1945.)

Parent-Young Interactions

The establishment of a spedes-bond by imprinting of young
has distinct advantages in adapting to a domestic environment
During the critical period of imprinhng, young can be separated
from parents in groups and raised by humans, thus alteraig t eir
sexual preferences little and sodalizing them to man. recocia
young have an earlier opportunity than altridal young or impnn
ing during the critical period. Acceptance of ahen young by some
spedes (allowing transfer of wild young and * a °P
tates successful establishment of a speaes un er o nthpr

tions. Recognition of young by spedes color because

spedfic cues) is less adaptive under th«e con 1 . ^ .

fLales might kill young of other mothers because of vanations in
color patterns or overt behavior per se.

""Tston n^h. distance and minimum ^isnipt^n -h the pm.

ence of human! fadlitates handling and reanng.

man early in life W )““ doe! imprinting to

Taming

14

THE dog; rrs domestication and behavior

cialization in wild versus domestic canids is discussed in detail in
Chapter 8. Genetic selection for high proximity tolerance and high
response threshold and flight distance to novel stimuli may have
been a consequence of selection for docility (see also Beylaev and
Trut, 1975).

Socialization plays a key role in both the initial phase of
domesticating a species and in maintaining domestic conditions
with each generation. An animal must be socialized to humans
without altering the sexual preferences of the species or causing
rejection of the animal by its own species. Constant handling at a
young age will socialize wild rats to humans, but the rats must be
handled throughout their lives to maintain this tractability (Richter,
1954). Similar "regression" in the absence of continued social con-
tact has been noted in some foxes and coyotes, although individual
variability is considerable. Wild adult animals may become
habituated (or tamed) to man, but only the more sociable and gre-
garious (e.g., wolves [Woolpy, 1968I) can become socialized (i.e.,
emotionally attached) when adult. When young, socialization is
usually very rapid but may not endure if that species is by nature
nongregarious or relatively asocial (see Chapter 8 for further obser-
vations).

Table III, modified from Fox (1968b), evaluates the various
stages involved in the socialization of animals to humans.

Food and Habitat Characteristics

"Ability to meet needs for food in man's environment or live
on the byproducts of his agriculture is essential for complete
domestication" (Hale, 1962). A flexible diet is adaptive under
domestic conditions.

Specialized habitat requirements are a disadvantage under
domestic conditions as confinement prevents searching for ideal
habitats. Thus limited agility is also adaptive.

BEHAVIORAL CHANGES INDUCED UNDER
DOMESTICATION

Eibl-Eibesfeldt (1970) views domestication as resulHng in generally
simpler behavior. Alterations in characteristic behavior patterns

Effects of Domestication

15

under domestication are the result of the various adaptations a

species must make to a given set of conditions. Such adaptations,
in many cases are learned, and modification occurs in the appeti-
tive sequences of behavior. An example of this might be a change
in the manner a domestic animal searches for food. Pigs might
learn that food is kept in a feeder and it can be obtained by lifting
the cover of the feeder rather than by searching for food in a field or
pen. New appetitive behavior leads to the species' characteristic
consummatory response, in this case, eating behavior characteris-
tic of pigs, as these patterns of behavior are highly stable.

Adjustments in behavior can also result from simply learning
to respond in characteristic ways to new stimuli associated with the
domestic environment. Stimulation for the release of milk in dairy
cows is frequently switched from manipulation of the udder by
calves to the rattle of cages and equipment at milking time (Ely and

Peterson, 1941). , . ,u

The effects of selective breeding are readily observed m the
more malleable behavioral traits. As has been discussed, most ^sic
motor patterns resist modification under domestic condihons. Cer-
tain of these patterns will combine through hybndizahon some-
times creating difficulties because of ambivalence in the reactions

of the hybrid. , , , j „

Under domestication, territorial patterns tend to
Those species with a greater flexibility in their soaa s . , ,
generally adapt better to the domestic environmen . a
points out, though, that some species have a . jg

vergent social structures." Some species whose soa ^ P. . ,
based on hierarchical structure in a wild habitat '

under domestic conditions. Changes also occur arranee-

habits of some species where reversion from pair-bond arrange-
ments to promiscuous mating takes place. frpauencv

Selecttve breeding shows a marked effect ‘he frequen^

and intensity of certain patterns of behavior ^ Apparent

response thresholds necessary to eUa‘ populaHons

elimination of certain of these patterns 1 This has been seen

is the result of an increase of these ^eshNds. Th. has
in experiments with a number “ SP ^t. It has been

experimentation has been done w^h . es^cated strains of
shown (Barnett and Stoddart, 1969) ‘hat domesnca

i6

the dog: its domestication and behavior

this species have higher thresholds for aggression in a conflict situ-
ation and are less suspicious of new objects, food, etc., as a result
of a higher threshold for avoidance behavior and neophobia (Bar-
nett, 1958). Keeler (1970) has shown that selective breeding for coat
color in color-phase foxes produces higher thresholds for behavior
associated with the fear response.*

Because the domestic environment removes many of the selec-
tion pressures of the natural habitat, behavior patterns adaptive to
competition for survival are not as adaptive under domestication.
Domestication does not remove responses of fright, aggression,
etc., from the repertoire of behavior of a species, and occasionally
such behavior patterns occur spontaneously as vacuum activities
(Fox, 1968b). More usually, however, without reinforcement and
no longer being of adaptive value, they may be easily eliminated
through selective breeding.

"Purely intra-specific selective breeding can lead to the de-
velopment of forms and behaviour patterns which are not only
nonadaptlve, but can even have adverse effects on species preser-
vation" (Lorenz, 1968). Selection for traits of economic advantage
to the breeder can result in disruption of the coordination of certain
patterns of behavior unless stability of these behavioral traits is
bred for simultaneously. This disruption has been seen in certain
strains of turkeys bred for extreme breast size. Some males exhibit
sexual-copulatory behavior toward piles of dirt, rarely mounting
females. Perpetuation of these strains is accomplished by artificial
insemination (Hale, 1962). Complete evaluation of many of these
behavioral disruptions is complicated by the lack of adequate in-
formation on the occurrence of such patterns in wild strains.

Threshold changes induced by domestication have already
been mentioned. Occasionally selection for certain behavioral traits
leads to undesirable changes in other patterns of behavior. Certain
strains of turkeys bred for hypersexuality have shown decreased
thresholds for imprinting to man, resulting in the display of sexual
behavior directed toward the caretaker!

Another maladaptive consequence of intense selective breed-

•In view of the marked differences within dog brecdsof individuals having different
coat colors, a dcuilcd study of the relationship between coal color and tempera-
ment IS needed.

Effects of Domestication

17

ing is the appearance of behavioral phenodeviants in hybrids
which are not apparently present in either parent strain. A certain
hybrid strain of turkey has been observed denuding many mem-
bers of the flock spontaneously. Deviant behavior has been seen in
mammals where in one case a hybrid strain of female mice is
known to eat specific digits from her offspring while cleaning
them. This phenomenon, according to Hale (1962), remains basi-
cally unexplained.

NEW ADAPTIVE PEAK ACHIEVED UNDER
DOMESTICATION

As man domesticates a species he is endeavouring to shift that
species from the adaptive peak representing the natural habitat
and ecological niche to a new adaptive peak representing the artifi-
cial, domestic habitat. The success of this shift is primarily depen-
dent on man's ability to maintain the reproductive fitness ot the
species. Many species have a number of characteristics favora e to
domesticahon, and the valley between peaks is a shaUower one
than the transition for species with fewer favorable charactenstics.
Occasionally, a species well adapted to the domestic environmen
may lose some reproductive fitness during selective ree mg or
specific specialized traits. This problem can be dea t wi y re
ing selection for the desired trait causing some regression o
trait, but allowing the reproductive fitness of the stram ‘o
a new gene pool to establish itself, and thus copu a ^ ® ,

new adaptive peak. Hybrids produced from crosses ^

mestic and wild strains of the same species frequen y

nations of characteristics unfavorable for succes

either habitat. Crossbreeding between two or considerable

domesticated animals (dogs or cattle) can hybrid

improvement in performance — a valuable as

vigor.

i8

THE dog; rrs domestication and behavior

The Effects of
Domestication on
Physiology

Changes in the environment alter the physiological needs of a
species, and corresponding adjustments of behavior occur. Thus, it
is important to consider physiological change as a basis for many of
the changes in behavior due to domestication.

The hypothalamus has been implicated in changes in the de-
gree of characteristic aggressiveness in wild and laboratory bred
rats. As Barnett and Stoddart (1969) have pointed out, laboratory
bred rats are much less prone to attack or threaten a strange rat in a
conflict situation than a wild rat which will frequently kill a
Stranger. This "killer instinct" has been "tamed" in wild strains by
the injection of methyl atropine into the hypothalamus and surfaced
in domesticated strains by a similar injection of various cholinomi-
metics. Rats are indicative of the general trend toward docility in
domesticated animals, and the possibility of hypothalamic change as
a cause of this is currently being researched.

Richter's (1954) work with the Norway rat has shown that
domesticated strains have a lower characteristic metabolic rate re-
sulting in a lower food and water intake per kilogram of body
weight. Coupled with this is a lower resistance to poisoning than is
seen in wild rats.

Nervous system changes accompany domestication. Changes
in the frontal poles are another factor in the docility seen in domes-
tic rats. Removal of the frontal poles causes domesticated rats to
lose much of the savageness of wild rats (Richter, 1954). The sen-
sitivity of various neural structures to hormones may change.
These specific changes cause changes in both the physiological and
behavioral thresholds of the strain (Hale, 1962).

Beylaev and Trut (1975), in selecting for docility in successive
generations of silver foxes, found changes in morphology, estrous
C)-cles, and adrenal responses to stress and ACTH administration.
They conclude that "selection for dodle, tractable behaviour leads
to the dramatic emergence of new forms (phenotypes) and to the

Effects of Domestication

19

destablilizaKon of ontogenesis manifested by the breakdown of
correlated systems (adrenal-pituitary, gonadal-pituitary) created
originally under stabilizing selection." They present an important
concept of destabilization where artificial selection can increase
phenotypic variance and alter the wild phenotype (structurally,
physiologically, and behaviorally) as exemplified by the domestica-
tion of animals.

Several of the general and theoretical issues pertaining to the
effects of domestication are examined experimentally, with first a
detailed analysis of the social behavior of wild versus domestic
canids.

II

Socio-Ecology of Wild
Canids: Environment
and Behavioral
Adaptation

Introduction

22

THE dog: its domestication and behavior

The Distribution of
Canids ,

The domesticated dog has many cousins which belong to the fam-
Uy Canidae (see Figure i). These include the coyote and wolf, vari-
ous species of fox and jackal, the lirdian dhole, and the Cape hunt-
ing dog (Staines, 1975). The doglike "foxes" of South America,
such as the maned "wolf," bush dog, small-eared dog, crab-eating
fox, and Culpeo are unlike any of the other canids, their unique-
ness probably being related to the great length of time that the
continent of South America was effectively cut off from the rest of
the world. In contrast, the jackals of Africa and Asia more closely
resemble the coyote of North America and, similarly, the pack-
hunting dhole of India resembles somewhat the Cape hunting
dog. These similarities may be interpreted as evolutionary paral-
lelisms and convergences. For example, the coyote and golden
jackal may not, in fact, be distantly related, but, instead, have come
to resemble each other because of their specialization to a particular
niche or life-style.

In the Northern hemisphere, the wolf is widely distributed
and may be regarded as the most successful of the pack hunters
(the other two pack hunters being the Cape hunting dog and the
dhole), being found as far north as the Arctic, as far south as
Mexico, and as far east as Asia. There were many races or sub-
species of wolf, most of which are now extinct because of man's
invasion and exploitation of their hunting ranges and systematic
destruction of predator species. Consequently many races of
wolves, superbly adapted to particular regions after generations of
evolution, are now extinct. Similarly, the red wolf {Canis nigerl
nifus), once abundant in Texas and Florida, is now virtually extinct.
A few specimens still remain in the Texas gulf area, many of which
have been crossbred with more tenacious coyotes, so that few, if
any, red wolves are left.

The other two species of pack hunting canids are also facing
possible extinction. Until it was realized that predation by Cape
hunting dogs was beneficial to quality and population control of
hoofed animals in Africa, the dogs were shot on sight. (In some
areas they are still shot for "disturbing" the herds of big game!) In

Socio-Ecology of Wild Canids

23

Vulpes spp.

Fennecus

Alopex

Urocyon

Otocyon

Nycfereufes

Cerdocyon

Atelocynus

Speofhos

Chrysocyon

Dusicyon

Cuon

Lycaon

C. lupus

/[ C. forniliaris

I C. rufus

\ aureus

mesomelas

odustus

C. latrons

Figure 1 . Schema of taxonomic relationships

based upon behavioral evidence ili.irrs and other

mon origin of uvif (C. lupus) and dog (C. • nnor to

uvtflike canids), probably differentiating as separ ./

mafi'5 domestication of the dog.

24

THE dog: its domestication and behavior

India, the dhole is in a similarly precarious position, as are many of
the carnivores (see Fox, 1978).

In contrast, the more solitary predators, such as the red fox
and coyote, are more adaptable to human intrusion and predation.
The elusive coyote numbers at least 40,000 in Kansas alone, while
in England, there is a high fox population in Tilbury docks, which
competes with the indigenous warehouse cats for rats and garbage.

The raccoonlike dog from Japan, which in the fall becomes
very fat and may hibernate in the winter, is another example of a
successful canid. Imported to Finland for commercial fur produc-
tion, it escaped from the ranches and is now wild in many Euro-
pean countries including Poland and Czechoslovakia. The dingo of
Australia, which closely resembles the smaller dogs of New Guinea
and Malaysia, was originally a domesticated dog brought into the
continent by early aboriginal settlers. It has since gone wild, be-
coming well established as the main predator because there were
no other placental carnivores with which to compete (Macintosh,

1975)-

Having briefly considered the distribution of canids, we will
now look at their behavior. The comparative analysis of social or-
ganization and communication signals can give us some insight
into the evolution of social behavior within a family of related
species. Similarly, comparative studies of the same species in dif-
ferent habitats show how such a species has adapted socially and
physically to a particular set of environmental conditions (e.g.,
distribution, abundance, and type of prey) and differences in ter-
rain (mountain, tundra, or open prairie).

Communication and
Social Organization

The canid family comprises an assembly of species that span the
entire spectrum from the relatively solitary red fox to the gregari-
ous wolf. Consequently, this family provides a valuable research
opportunity for studying the evolution of communication
mechanisms in related species differing in social behavior and in

Socio-Ecology of Wild Canids

25

the sodo-ecological patterns or life-styles that they have evolved.
The relatively solitary canids such as the red fox have a simple
repertoire of visual signals — tail and body positions and facial ex-
pressions. Similar basic signals which serve to increase, decrease,
or maintain a certain social distance or proximity are recognizable
in the more social canids, but in these species, the signals are more
variable and are subtly graded in intensity. Greater complexity of
signals and, therefore, greater message carrying potential is also
afforded by rapid successive and simultaneous combinations of
these signals. Thus, the wolf and domesticated dog can give suc-
cessive alternate signals of submission and defensive aggression or
can combine, simultaneously, signals of submission and greeting.
Very similar signals are present in the less social canids, but they
often appear at a "typical intensity" and lack these more subtle
intensity shifts. The threat gape of a red fox is very much an all or
nothing signal, while the wolf has a greater repertoire of threat
signaling (Fox, 1971b).

These generalizations may point to the fact that with increas-
ing sociability and sustained proximity, a strong selection priority
was placed upon the development of more sophisticated visual
signals.

Marler and Hamilton (1966) have pointed out that the more
solitary primates have highly stereotyped calls, while the more
sociable species have more subtle and variable vocal patterns. A
similar generalization may be applied to the canids, the more soli-
tary of which have elaborate and stereotyped calls which are used
especially during the breeding season to locate a mate and to drive
off rivals. The more solitary the spedes (which also tend to be
nocturnal), the greater is the reliance and priority upon auditory
communication because visual signals are ineffectual when indi-
viduals are separated by some distance. Wolves and coyotes also
have distance-communication signals; their hoivis, interspersed
with barks and yips, serving to bring companions together and
possibly to inform rival individuals or packs of their presence. The
more solitary canids apparently have not evolved a howl in their
vocal repertoire, so that the howl seems to be correlated with more
complex social interaction (see also Chapter 4). Wolves and Cape
hunting dogs often engage in a mutual greeting and singing (or
howling) ceremony before they go out to hunt; such group K-

26

THE DOG: ITS DOMESTICATIOM AND BEHAVIOR

haviormay serve as a pep rally, bringing all individuals to the same
degree of excitement.

The more solitary canids also seem to rely a great deal upon
olfactory communication. Odors (pheromones), when deposited
on the ground in urine or feces or against particular objects or
"scent posts," tend to persist for a long time. The message is,
therefore, a rnuch more permanent kind of signal and is effective
long after the animal has moved on. Social canids mark out their
territories and, like howling, it may serve to inform others of their
presence and of their movements in the hunting range which others
might use. It would seem logical that the less social canids rely
more upon olfactory communication than the more gregarious
species, and this is partly supported by the simple observation that
the more solitary species smell much stronger than the more social
ones. They are more likely to urinate or defecate on a novel odor (of
carrion or deer musk), while the more social canids will often roll
on such odors, their reward possibly being a lot of social investiga-
tion when they join up with their companions later (Fox, personal
observations). Olfactory, as well as auditory, communication would
be essential in order for the male to locate a receptive female during
the breeding season. In the more social canids, where the receptive
female is always with her mate or always with the pack, the selection
pressures for effective olfactory and auditory commurucation would
be relaxed.

Social-Ecological

Interactions

The communication patterns are related to the type of social or-
ganization that the various canids have evolved, which in turn
reflects the ecological factors to which each species has adapted or
has been shaped by. These ecological factors include primarily the
species size, annual and seasonal distrlbutiorv, and abundance of
prey. The type of terrain, climate, and seasonal variation in tem-
perature are also significant, as well as the presence of other pred-

Socio-Ecology of Wild Canids

27

GINOrr^lc/PHtNOTYPtC CHARACTERISTICS

Bod/ Size I

Timing of ^Motu/ity at birth^Porent-infont *
Breeding Seoson^ Utter size

^Communicotion
Sociof Distance ^Signois

^Temperament

(proximity tolerance)

Figure 2. Schema of interrelated genetic and environmental var-
iables which determine genotxfpicfphenotypic characteristics in the
socio-ecology of canids (from Fox, 1975). Human intervention, as
in domestication, can clearly influence these interrelationships at
almost every interphase, leading to increasing genotypic destabiliza-
tion and phenotypic variance in such variables as proximity tol-
erance, temperament, body size, and reproduction.

ators of greater or lesser size which occupy competitive or non-
competitive niches in the same habitat.

Competitive ruches are rare and when competition between
predators occurs, the ecological balance may be disturbed. Where
there is an abundance of many different prey species ranging
greatly in size, a greater diversity of predators can be supported.
Thus, wolf, coyote, and red fox, or Cape hunting dog, golden
jackal, and bat-eared fox may be found in the same habitat, each in
its own particular niche and each occasionally scavenging on the
remains of a kill made by a larger predator/ Indirectly, one pred-
ator may assist another in the complex food chain by regulating
population growth and preventing starvation. For example,
smaller predators that specialize on small herbivorous mammals
(rabbits and rodents) prevent the latter from becoming too abun-
dant which, if they did, would lead to overgrazing which could

’Interestingly, and tragically, the extermination of the wolf in .

resulted in a decline in the wolverine, Arctic fox, owl, and raven, an

that relied greatly on the remains of wolf kiUs as their pnme source of food

28

THE dog; its domestication and behavior

cause infertility, starvation, and death of larger herbivores such as
deer. The larger predators that depend on the latter would then
suffer.

The equilibrium maintained between prey and predator popu-
lations is remarkable, the complexities of which are only just be-
ginning to be understood. Wolf packs in Alaska and Isle Royale
have remained more or less the same size over the past 30 years,
and even in years when prey is abundant, the pack size does not
suddenly increase (Haber, personal communication). Many of the
factors which regulate pack size remain to be identified, but some
have been detailed (Fox, 1973 and Fox effl/., 1974). In most canids, the
breeding season is timed so that the birth of young will coincide
with the birth of prey species, at which time food is most plentiful.
The prey tend to give birth all at the same lime, such reproductive
and parturient synchrony possibly enhancing survival because the
predators would eat their fill and be satisfied. If the birth of young
were not synchronized but instead extended over a longer period,
predators would take a far greater toll on the newborn before the
latter were able to fend for themselves.

Social Organization

Theie are three basic patterns of social organization in the Canidae.
In Type I, such as the red fox, only a temporary bond is formed
during the breeding season between male and female (Burrows,
1968). The male may stay with the female and assist in providing
food for the young, but the Utter is usually deserted by the parents
around 4-5 months of age. No stable hierarchy or peck order de-
velops in the litter, so that, in the absence of strong social bonds
and other group-cohesive forces such as leader-follower relation-
ships, the litter breaks up and individuals go their separate ways as
solitary hunters.

In the Type II canid, such as the coyote (Gier, 1975) and golden
jackal (Van Lawick and Van Lawick Goodall, 1971), there is a more
or less permanent pair-bond between male and female, and their
offspring may remain with them until the following breeding sea-
son. They may then be driven out of the parents' territory, but
occasionally when there is an abundance of food, and therefore

Socio-Ecology of Wild Canids

29

reduced competition, some yearlings may stay and even assist the
parents in tending for the next batch of offspring. Such occurrences
are rare. More usually the litter disperses, and each member seeks
its own fortune outside of its natal territory. How early familial
relationships influence later social interactions between Type II
canids sharing the same hunting range remain to be evaluated.

The most detailed studies of the sociai dynamics in Type III
canids have been done on the wolf (Mech, 1970). There is stiil some
doubt about the origin of the pack, but the general consensus is
that the pack consists of related individuals of various ages. There
is a binding leader-follower relationship between subordinate
wolves and the lead or alpha wolf. A dominance order is seen
among the females and the males, the alpha, or most dominant
male, serving not only as a leader and decision-maker but also as a
"policeman” who often intervenes to settle disputes that flare up in
the ranks. There are pack rituals that may serve to maintain and
reinforce social bonds; subordinates will affectionately and submis-

sively greet and mob the leader, while the leader may present
the pack with some token food object. Pack members will assist a
mated pair in feeding the young and in guarding them when the
pack is away on a hunt. As the young mature, they may be re-
cruited into the adult pack and essentially fill vacant places af-
forded by the death or departure of others. If such vac^cies are
not available and food is scarce, mortalities will be high. Occasion-
ally such a large pack may spUt up and the yourig, their parents,
and one or two other adults may move into a different hunting
range or remain in the same locale while another segment of the
pack moves out. Such emigrations would on^ be possib e where
neighboring packs are not in posession of adjacent temtory an
hunbng range. The factors which reflate pack size over genera-
tions of offspring remain to be idenhfied. It is known at
every adult in the pack does not breed and that there is a kind of
socid control of breeding.* Usually
breeds; she subordinates other femaks and

show any overt sexual behavior to the mate of he choice (nhich
may or may not be the alpha male) or to other males.

•Group vocrlizations in wolves and coyotes may have some
cpidcictic function) to regulate reproduction

Figure yhlerspcdes differmcc in mlraspecirw aggression
Coyote cabs are extremely aggressive while Jlf ente of the a
age engage m sustained bouts of play. Tacing page: At a lal

Socio-Ecology of Wild Canids

31

32

THE dog: its domestication and behavior

Temperament and
Behavior

What determines whether or not a young fox or wolf cub will live a
solitary or group-oriented life? It seems that much is determined by
the basic temperament of the species, as well as by the presence or
absence of the group-cohesive forces alluded to earlier. We might
postulate that a socal drive independent of a sexual drive in the
Type H and Type HI canids keeps individuals together outside of
the breeding season. Recent research on young wolf, coyote, and
red fox cubs supports the notion that temperament is also a major
determining factor (Fox, 1975). Young foxes, from an early age,
show great proximity intolerance; each one is an individual and
each is invariably very confident, inquisitive, and highly efficient at
killing prey. In wolf litters, there is a much greater range in indi-
vidual temperaments, ranging from the shy, timid type (who are
often too afraid to attack small prey) who tend to be the subordi-
nates to the confident and outgoing type. The latter are the domi-
nant wolves who are very inquisitive and kill prey at an early age.
They also act as leaders or initiators for their less outgoing litter-
mates. Such a spectrum of individual differences, coupled with
group-coordinated behavior, social facilitation of actions, and a
strong tendency to follow the leader, insures pack formation.

There may be a strong selection primarily for such hetero-
geneity of temperaments (behavioral polymorphism) within wolf
Utters, while in the Type I canid, selection has been for a greater
homogeneity of temperament. Litters of coyotes so far studied,
as would be anticipated, lie somewhere between the red fox
and the wolf in that they show considerable proximity intolerance
and aggression toward each other, have little group-coordinated
^havior and leader-follower tendencies, but are capable of form-
ing a fairly stable dominance hierarchy.

Socio-Ecology of Wild Canids

33

Development of
Aggression and Social
Bonds

Wolf cubs begin to play around 3 weeks of age, and their bites,
unlike coyotes and foxes, are gentle and controlled. They rarely
fight — usually to establish dominance relationships around 5-8
weeks of age, especially in those individuals having similar outgo-
ing temperaments. In contrast, the fox species, and coyotes and
jackals, are very aggressive, and fighting often occurs before they
actually engage in play. At 3 weeks of age the bite is not as inhib-
ited as in the wolf, and they seem to have to leam to control it (Fox,
1971b).

When first put together, domestic dog pups that have been
raised apart from each other in social isolation until 12 weeks of age
play for only brief periods. Play bouts are broken by one biting its
partner too hard. After 3-4 days, though, play bouts are more
sustained because they have now learned to control the bite.
Therefore, social experience, as well as genetic influences, contrib-

utes to the control of bite intensity.

Social relationships are based primarily upon who is domi-
nant, and dominance relationships are usually firmly established in
coyotes and jackals by 4-5 weeks of age. By this age, they are more
likely to play, the subordinates often showing more play-solicitmg
than their more aggressive littermales.

Foxes, although very playful by this age, do not forrn close
affecHonal Hes with each other, and play often ends m a fight. Fox
cubs rarely engage in group activiHes-each one is a confident indi-
vidualist. Their egocentric temperaments (high '

erance) and loose social ties in infanc>- suit them ivell to
tary life; by 5 months the litter splits up each one going it.s ou n
separate way The parents usually leave the young and at this time
the bond bL-een the parents is also broken. [Occasionally the
bond may break after mating, and the vixen raises the cubs herself

‘’’“"in conirat/tL the fox, a litter of wolf cubs has a greater range of

34

the dog: its domestication and behavior

temperaments. One or two are dominant and outgoing leaders,
others are followers, and some are very timid and dependent. Es-
kimo hunters state that some wolves never kill prey— perhaps
these belong to the latter category. Research on several litters of
wolf cubs at the Naval Arctic Research Laboratory at Point Barrow,
Alaska has confirmed this polymorphic heterogeneity of tempera-
ments in wolf litters (in contrast to the monomorphic temperament
of more homogeneous fox litters [Fox, 1975]). This phenomenon in
wolf litters may facilitate pack formation; all cannot be leaders and
there must be followers. A greater range of temperaments would
facilitate division of labor, of activities or roles in a social group
comprised of individuals of different rank, sex, age, and having
different temperaments and allegiances. Dr. David Mech (personal
communication) has been studying several packs of wolves marked
with radio collars in Minnesota and has also tracked several loner
wolves. These are not all old outcasts or low-ranking youngsters;
some seem to be young dominant wolves that simply cannot inte-
grate with the pack in which there is a firmly entrenched leader.

Fox and Andrews (1973) have recently found significant
physiological differences, as well as behavioral differences, in wolf
cubs of high, low, and Intermediate rank. High-ranking wolf cubs
and yearlings have a greater sympathetic tone (higher resting heart
rates) and are both behaviorally and physiologically more active
than subordinates. These physiological differences discovered in
wolf cubs may be the basis for differences in temperament or per-
sonality. It remains to be seen if, when a wolf experiences a change
in rank, there is concorrritant physiological change. For example,
would a low-rankmg cub with a low resting heart rate have a
higher heart rate if it experienced an elevation in rank? Candland d
al . (1970) did find that heart rate changes with a change in social
rank in groups of chickens and squirrel monkeys.

Foxes together in captivity show little group-coordinated activ-
ity and do not follow a leader like the wolf. This innate tendency to
follow may be a major reason why wolf cubs stay in a pack.
Another reason may be the maintained responsiveness of wolf
cubs to their parents (and vice versa). Fox and coyote cubs show a
decreasing responsiveness to their parents as they mature, while
wolf cubs continue to be highly responsive. A transference may
occur where the young wolf reacts to the leader (which may or may

Socio-Ecology of Wild Canids

35

not be a parent) in the same way in which it behaved toward the
parent as a cub.

The temperaments of young foxes and wolves seem well fitted
to their particular life-styles. If foxes were to hunt in packs, one
mouse or rabbit would not go far. Since their major prey is small
and widely dispersed, their temperament and solitary method of
predation are superbly adapted to their particular ecological niche.
Wolves, in contrast, cooperate in hunting and are able to transcend
their body size limitation by cooperating to secure prey many times
larger than themselves. Wolves that prey mainly on deer tend to be
smaller in size and in smaller packs than other races that hunt
moose, for example. When food is scarce, the wolf pack will tem-
porarily split up, and in some areas where food is scarce, as in
Mexico, wolves are rarely seen in packs. In Italy, they do not hunt
in packs but are mainly solitary village scavengers (Zimen, per-
sonal communication). Such regional ecological differences have
contributed to the evolution of various races of wolves.

The coyotes represent a canid type intermediate between the
fox and the wolf. They neither form permanent packs nor live a
relatively solitary existence but instead tend to maintain a perma-
nent pair-bond. Litters of coyotes show a greater range of temper-
ament than the red fox but less variants than in the wolf, i.e.,
oligomorphic. They do not disperse until around ii months of age
(i.e., around the time that the parents come into breeding condi-
tion). Again, their temperament, greater proximity intolerance for
each other, and relative lack of group-coordinated and leader-
focused behavior compared to the wolf accords well with their
social ecology. They hunt mainly small prey (rabbits, small ungu-
lates) and usuaUy cooperate in pairs. Occasionally when there is an
abundance of food, the young, or some of the ^ter, may s‘ay ™th
the parents and help them rear another lifter. is is a
type^of social behavior more typical of the wolf and has also been
reported by the Van Lawicks (1971) i" the golden
and jackals under certain ecological conditmns may there ore show
some of the social patterns of wolves. This
capacities are more flexible than in the re y' L ' -Uservi-
food is abundant, still tends torvard a solitary li e. “^.va

tions warrant further experimenlaHon to
social (experienHal) determinants of behavioral adaptation

36

THE dog: rrs domestication and behavior

ecology. It would seem that the fox is genetically limited to one par-
ticular life-style and ecological niche, while the coyote and wolf are
more flexible. Kummer (1971), in his field studies of various races
and subspecies of baboons, poses the same question.

In conclusion, the red fox, coyote, and wolf represent three
basic canid types; Type I, solitary, with temporary pair-bond; Type
II, permanent pair-bond, with longer interaction with offspring;
Type III, pack-forming, with strong allegiances and pack affilia-
tions and with dominance hierarchies in both male and female
ranks. Under certain sets of ecological conditions, the Type II canid
may show some of the patterns of Type III, while Type III may
adopt a Type II pattern as exemplified by the Mexican and Italian
wolves.

It is not too premature at this stage to hypothesize that temp-
erament seems to be strongly correlated with the ecology and life-
style of the individual and that past social and ecological factors are
responsible for the selection and continuation of temperament
types in the various wild canid species. Relaxation of such selection
factors in domestication might therefore lead to greater variations
in species-typical temperaments;’ some breeds of dog resemble
foxes, coyotes, or alpha wolves (e.g., nongregarious, aggressive
terriers), while others are more like middle- and low-ranking
"omega" wolves (e.g., gregarious beagles, often lacking any social
dominance hierarchy).

Domestication

The origin(s) of the domesticated dog and the various breeds is
unknown (Fox and Bekoff, 1975). There may be wolf and even
coyote and jackal ancestry (or perhaps these bloodlines are only in
some breeds)— these three wild species all will produce fertile hy-
brids when bred with domesticated dogs (Gray, 1954, and Koleno-
sky, 1971). These may have been added to the genetic diversity
wrought by 10,000 years of domesticating a dingolike dog ancestor.
Certainly the range, variation of size, temperament, and specialist

•Which U..ids to Lon-nz's now ui:.inlcd view of separate j.ick.il and wolf ancestry for
”Dnc*m.innish" and greganous breeds of dog, respectively.

Socio-Ecology of Wild Canids

37

abilities of the various contemporary breeds is an incredible
example of what can be done under intensive artificial selection. To
see a naive sheep dog pup playing with a companion and showing
all the basic actions of sheepherding, including blocking, turning,
and driving is a dramatic case of genetic engineering. Certain
behaviors seem to be at a lower threshold and are more easily
elicited in some breeds than in others (e.g., herding, retrieving),
and they are therefore easier to reinforce in training the dog. Other
actions have a greater amplitude or intensity in some breeds; the
pointing of bird dogs, for example, is more exaggerated than in
terriers, while the action to attack or chase has been truncated in
pointers (at the juncture of “orientation" in the temporal se-
quence). Terriers show a strong tendency to seek out (i.e.,
appetitive drive) suitable objects to chase or attack, whereas a
sheep dog may guard the same object. These are examples of
instinct enhancement which has led to various breeds of dog

fulfilling particular roles for man.

The dog differs greatly from other wild canids, not in overt
behavior per se, but in its vocal repertoire (see Chapter 4) and also
in its sexual and social behavior. In order to successfully breed
many dogs, the trait of monogamy or specific mate preference
evident in wild canids has been almost eliminated. A desirable stud
dog that is only attracted to one particular female would be of little
value, and vice versa. Some dogs sHIl show this ancestral trait
(Beach and LeBoeuf, 1967); female beagles and Boston temers, for
example, if given the opportunity, have been known to accept only

certain males. . . ,

Wild canids only have one breeding season per year, which is
Hmed so that there will be an abundance of prey for the young
Domestic dogs, with the excepHon of the Basenp (which has a
photoperiodically controlled single annual estrus IScott and F ,
1965]), have two and sometimes three heats per year, ‘'•'’d h s trait
is inherited. Artificial selecHon for this trad y/"]

hance production. ^;°“|his js^inLTd the case in the

domTslrc d^g T^t: nstxJ;. matS - early as 6 months, in
uuiiie&iiL uug. ati Here a combination of

contrast to 2 years in the woU and coy invnivpd Male

genetic selection and improved nutntion "’fy ^ J
wild canids produce little or no spe^ Although th,?

season, while domesticated dogs are always potent. Although th,.

38

THE dog: its domestication and behavior

again enhances productivity, it can become a serious social prob-
lem (see Chapter u). Dogs, then, selectively bred for utilitanan
purposes are sexually promiscuous, precocious, and prepotent,
and such activities can be afforded since the natural ecological re-
straints no longer operate.

Experiences with hand-raised canids have shown that coyotes
and wolves, especially, become increasingly fearful of strange ob-
jects or to a change in their familiar environment around 4“5
months of age. In the wild, this environmental fear may be a con-
sequence of exposure learning (or imprinting to the home range).
It would be highly adaptive for a young animal to recognize any
change in its familiar territory which could mean danger. In con-
trast, the dog does not normally manifest such behavior, presum-
ably because in the protective domestic environment which is con-
stantly changing, such behavior would no longer be adaptive and
would make such dogs difficult to handle as well as interfering
with performance. An analogous fear of strangers and the capacity
to develop sodal relationships later in life will be discussed in
Chapter 8.

Wolf cubs show a greater tendency than coyotes to maintain
the infant attachment in that they invariably remain bonded to
their human "parent" throughout life. Wolves also seem to have a
greater capacity to develop secondary social relationships with
strange people later in life than other wild canids, but this social
potential is still much less than in the dog. The parent bond in the
wolf may be transmuted to a leader bond as the wolf cub matures.
A similar transition occurs under normal conditions in the relation-
ship between a dog and its master as the former matures. Less
gregarious carnivores, such as foxes, which are relatively solitary in
the wild and break away from their parents, also will grow away
from their human foster parents; the infant bond is not often trans-
ferred into a social bond (leader-follower relationship) with ma-
turity.

Domestic dogs may well have originated from a packtype an-
cestor, since their socialization patterns are closer to the Type III
canid than the Type I or II forms. Studies of feral urban and rural
dogs (see Chapter 3) reveal their capacity to form packs under
optimal environmental conditions, again supporting the view of a
Type III ancestry. Because of suboptimal environmental condi-

Socio-Ecology of Wild Canids

39

tions, pack formation is not possible for the dingo. In this species,
presumably, unidirectional selection over generations of feral liv-
ing has resulted in a temperament and social organization of the
Type II category (Corbett and Newsome, 1975). Dogs, like wolves,
also respond to a leader in adulthood, and this tendency is easily
transferred into the home environment where the dog becomes a
part of the household "pack" and is ideally subordinate to the
master of the house. But this is not always the case, and some of
the problems of socially maladjusted delinquent dogs are discussed

in Chapter 11.

In summary, a lot can be gleaned from studies of wild canids
in attempting to elucidate what effects domestication has had on
the dog. This brings us to an important variable, namely, the pro-
cess of socialization and critical and sensitive periods in develop-
ment, Some analogies will be subsequently made between the in-
terpersonal relationships of pet and owner and child and parent
(e.g., overdependent, overpermissive) which may give rise to a
variety of very similar emotional disorders in both dog and child.

Behavior patterns and communication signals in these canids
are inherited (in that they develop independent of social expen-
ence early in Ufe). Evidence has been presented here to support the
notion that the inheritance of certain temperament charactenstics
or traits (which are most adaptive to a particular life-style or ecolog-
ical niche) determines whether the animal will be sohtary or gre-
garious as an adult. Socialization during early e a so con “
significantly to the formation of family bonds. In other words, by
applying pLsonality theory and developmental
ethology Ld ecology of canids, a more complete picmre of their
behavifr in relation to the environment can be drawn A know,
edge of the natural history, social behavior and development o
each species is an essential prerequisite, and lUs unfortunate tha
only a few studies have so far been comp e e . pnvi-

In conclusion then, the truism that the f Xck

ronment is confirmed in terms of to a particu-

determine sociability, temperament, an a ap ty selection
lar prey-predator domesHc environ-

pressures under “"dihons of rap parricular physical

rL^tra :n7;:y:hScaSshasoccum^^ in the domesHcated

40

THE dog: its domestication and behavior

dog. To understand what changes have occurred in the behavior
and psychology of Canis familiaris, the variables of human influence
must be considered. These are discussed subsequently in Chapter
u. At this stage, some preliminary comparisons between wild and
domesticated canids are made in relation to undesirable wild traits
and desirable traits associated with sexuality, reproduction, and
socialization (see Table I).

Summary

Differences in behavior, including patterns of communication, so-
cial organization, interpersonal relationships, and hunting be-
havior in various canids are described. Also species differences in
the development of aggression and the significance of individual
differences in temperament are detailed. These behavioral dif-
ferences between members of the canid family reveal evolutionary
changes related to adaptation to a particular set of environmental
conditions. Eco-spedaUzation, or socio-ecological adaptation, to a
particular niche is manifest by a particular life-style and typologically
distinct array of behavioral characteristics (temperament). These
findings are relevant not only to understanding the evolution of
ecologically adaptive behavior patterns, but they also provide a
basis for interpreting some of the behavioral consequences of
domestication in dogs, where both genetic and environmental
changes have been effected by man.

Ill

Behavior and Ecology
of an Urban Feral Dog
Pack

Introduction

Much research has been done on the behavior geneHcs, ontogeny/
and ethology of the domestic ‘JogCams/amdians (Scott and FuUer,
1965; Fox, i%a, r97«). Most ethological th '

taken in capHve conditions, and the question anses as o what the
noi^al envLnment for studying the domeshc dog reaUy is^ Afer
approximately 10,000 years of domestication, the dog shU retains
th^e capacity to become feral. 'Boreas many dogs are conf^ne^ in
the home, others have free access to the

or urban. These are designated as elves

from those feral dogs that have no hom^e and^support^he^m^^

Independent of human ^^^n commer^l dog food

though, people will put assortment of feral and

for what ‘hey f and Nesbitt (1975) have studied

free-roaming ‘Jeg®’ environments, respectively,

both ecotypes in urban a ter focuses on the behavioral

The study d-^cussed m this ^

ecology of a tno of feral dog P L^uis Missouri, a city of

in an economically depressed area of Sb^Lo^^^ .

some 600,000 inhabitants, b

""®?n several areas of the city, feral dogs were

THE dog: its domestication and behavior

Figure la. Typical free-roaming feral urban dogs: (A) using avail-
able environmental resources, (B, C, and D) showing varying de-
grees of emaciation and disease, (D) being in a terminal state.

seen, many of which, by virtue of their condition and behavior,
were easily distinguished from free-roaming dogs. The latter were
usually in good physical condition, tolerated proximity of strangers
up to 4.5-6 m (15-20 ft), and when approached further would run-
off and seek refuge in their open backyards which were usually no
more than 91 m (no yd) away (see Figures la and ib). From the
security of their presumed home sites they would bark at the inves-
tigators. Feral dogs showed a much greater flight distance, were
more elusive and harder to follow. Sometimes they were flushed
out of abandoned buildings which showed signs of canid occu-
pancy (feces, shed hair on old mattresses and carpeting, chewed
food cans and cartons and toys, such as a chewed ball or stick).
Putrefying and mummified carcasses of dogs were also found in
some of these buildings. The physical signs characteristic of many
feral dogs were emadation and skin lesions (probably mange);
some were in such poor condition that they were unable to run

Behavior and Ecology of an Urban Feral Dog

43

Figure lb. Tree-roaming house dogs: M and C)

bage, and (B) leaving home territory for early m g f . ■

and social interaction in neighborhood

44

the dog: its domestication and behavior

away when first located. One was seen in the company of three
free-roaming dogs and over a period of 6 weeks became weaker,
more emaciated, and the skin lesions more extensive until it even-
tually succumbed. Of course many of the dogs initially seen could
not be classified into feral or free-roaming until we concentrated
the study on one particular neighborhood. The neighborhood cho-
sen was determined by the fortuitous discovery of a group of three
feral dogs that were spotted one morning and followed into an
abandoned house in which dog signs (feces, hair, play objects)
indicated that they had been there for several months.

Materials and
Methods

The study began in late March, 1973, becoming most concentrated
from May through July, with sporadic observations through Feb-
ruary, 1974. Most observations were done from an automobile
since the subjects were wary of pedestrians, and any attempt to
follow them on foot evoked the flight response. After 2 weeks, the
three feral dogs would allow closer proximity while the observers
remained in the car and would walk within 0.6 m (2 ft) of the car
when we were parked. They soon became habituated to our follow-
ing them provided we kept the car at a distance of 6-9 m (20-30 ft).

The pack, or trio, consisted of one tan shorthaired female
mongrel (f) approximately 16 kg (35 lb) weight. She had elongated
teats, a sign of repeated litters, and she was probably the oldest
dog of the trio. The other two dogs were male, one a yellow Ger-
man shepherd (Alsatian) (Y), weighing approximately 32 kg (70
lb), and the other a large mongrel (X), with medium-long brindle
hair, weighing approximately 29 kg (65 lb). The latter had a perma-
nent hip injury, causing intermittent lameness in the right hind
leg. All of the dogs were obviously mature (see Figure 2). All were
in poor physical condition at the inception of the study, the
shepherd being the most healthy of the three. The other two had
extensive skin lesions, probably mange infestations, but by July, all

Behavior and Ecology of an Urban Feral Dog

45

three were dear and their general condition much improved (pos-
sibly a benefidal effect of sunlight). Fecal analysis revealed heavy
infestations of hookworm {Ancylostoma and Uncinaria) and whip-
worm {Trichiuris). Although tempted, the observers provided no
food or medication and in no way interacted with these animals

during the course of the study.

Two, and often three, observers followed the animals on 4- to
6-hr shifts. A total of 90 hr of observation were completed, cover-
ing all periods of the day and night. The following data were col-
lected:

1. Daily Cycle of Activity. The time of day when active in the
area of study, either foraging in alleys, traveling, resting in t e
open, or hunting in the park was recorded. Much time dunng
dayUght hours was spent in one of two abandoned house^Later m
the summer the cool basement of one was Referred. TOe ot er
house, an upstairs apartment, had a conveniently si a e sma
dog door cut into the back door; the leather flap falcated when
one or more of the dogs was in or out of the house by being tucke

■"°2°“population Activity Index. Head counts were made at

various hours of day and night of free-roaming og could

of the trio's home range. This provided an

be compared with the acHvity patterns of 'he ^ Pa^l^ '^^ich

might, for various reasons (see ^scussiot), e movement

3. Movements and Home Range. The range and movement

patterns of the trio were plotted on a survey P , context

4. Marking Behavior. The anther m^ember of the

(presence of strange dog or marking aft Crraoinc after

Wo had marked, i e., marking over) were recorded. Scraping

marking was also noted. . |cad at

5. Leadership. As the trio traveled, the one to take

intersections and road crossings was ^^,cre made as

6 . Miscellaneous Social Interactions. NoUations ^

to the frequency and type of "’‘^.”‘^*’‘’')^j‘hcirhood residents and
riveen them and other free-roaming neighborhood

feral dogs that occasionally came ■ openness and sta-

7. Pack Cohesion (Fission and Fusi | . among the three

bility of the group of interindividual relationsh p.

Figure 2. (a) Y leaving "flaphauee"; (b) F and X following him;
(c) Y from basement of seeand house; (d) trio crossine

busy highway at S:is a.m. after park "hunt". Facing page-

c and/). /cma/c F marAs in part, X marSLr; am

Ig) tno resting in the open near second house. ,2:30 a.m.

Behavior and Ecology of an Urban Feral Dog

47

dogs were determined by (a) recording often other

member left the group and for how long, and (b)

dogs joined the trio and for how long. engaged primarily

8. Hunting Behavior. In the park, the g g S ^ ^

in chasing squLels. The number of observed chases was recorded

and their hunting strategy s^died

9. Other Dog Groups. In addition to We n

quency of occurrence of dogs egariousness of free-

corded. This would give an index of th gr g lead to

roaming dogs and would possibly disclose what factors

ssss™ «... i"

.973 10 lesl Ihe poss.bai.y of dituiol go to the parh- A‘

m the early part of this study the dog u^mes 1-2 hr earlier

this Hme, they would return to one of their homes

than during the summer schedule.

48

THE dog: its domestication and behavior

Results

ACTIVITY

In late spring, the trio emerged from their sleeping quarters (usu-
ally the second floor flaphouse, i.e., the house with the leather flap
covering the cut in the door) between 11:15 p.m. and 12:00 a.m. and
remained active until 6:30-7:00 a.m., when they would retire until
the following evening. During the summer, they retired later —
between 7:30 and 9:00 a.m., usually to the cool basement of the
second house, or they would move to this basement from the
flaphouse sometime during the day as the ambient temperature
rose up to 30°C (86°F). The later returns to shelter in the morn-
ing always followed a i-iVi hr period hunting squirrels in the park,
an activity which was not evident earlier in the study. A kill of
squinels was never observed. Activity was greatest then from
around midnight to early morning, during which time the ambient
temperature averaged some 7° less than during the day. A lower
temperature in the evening and early hours of the morning would
be anticipated but the buildings and pavements of this humid city
give off much absorbed day heat during the night.

The night and early morning activity was interspersed with
rest periods of varying duration (see Figure 2); the trio never re-
turned to the cover of its houses during this time (even during
heavy rain), preferring to sleep out in the open, on lawns or
porches of occupied houses. When active, they would systemati-
cally forage for food in the three alleys and in backyards which
opened onto these alleys (see Figure 3). Interactions with other
dogs occurred during these times and also in the park, and these
will be described later.

Interestingly, the head counts of free-roaming dogs which
gave a good index of the activity of other dogs in the neighborhood
did not follow the pattern of the trio during the day and early
evening (see Figure 4). Whereas the trio were under cover during
the day, free-roaming dogs showed two peaks of activity —
between 9:00 and 10:00 a.m. and around 7:00 p.m. The dramatic
decline in head counts between 6:00 and 9:00 a.m. corresponds

SEEN PER 30 MINUTES

THE dog: its domestication and behavior

50

Figure 4. Activity index of free-roaming dogs in study area
{number counted over a Vi-'hr period). The nctivity of the feral trio
is superimposed for comparison. The latter show no late morning
(8:oO'9:oo o.m.) or early evening (7:00-9:00 p.m.) och'oify peflfcs
and are most active during the early hours of the morning. Data
averaged from 90 hrs obseruoHons.

with many of the free-roaming dogs returning home, presumably
to be fed when their owners arise. The evening peak also correlates
with human activity; the streets are full of diildren playing and
adults conversing on their porch steps. The feral trio was never
seen out at this time, and it may be concluded that they were
avoiding human contact. Most dogs avoided the hottest time of the
day between 2:00 and 4:00 p.m.; free-roaming dogs, like the feral
trio, were most active during the early hours of the morning,
shortly before sunrise (Figure 2),

Behavior and Ecology of an Urban Feral Dog

51

MOVEMENTS AND HOME RANGE

The trio's main area was located within a low income, primarily
abandoned, residential area of about four city streets of 427.5 m
(1402.5 ft) long in each direction and 33-5 ^ representing an

available surface area of about 5.7 ha (hectares) (14.2 acres) exclu-
sive of inner buildings. In the mornings, they often used much of
an adjacent 55.3 -ha (136.6 acres) park (477.6 m x 1156.7 m) or a total
home range of 61.0 ha (150.7 acres). This area is considerably larger
than the area used by a group of two feral dogs observed in
more, Maryland (Beck, 1973) possibly indicating a less favorable
habitat, i.e., food availability. Home range for many animals is
probably influenced by food availability (McNab, 1963)*^^^ oxes,
Vulpes fulva, were observed to have smaller ranges in °

greater ecological diversity than those in suboptimal habitat (Abies,
1969). Feral dogs in rural Alabama have home ranges grea er an
1,000 ha (Scott and Causey, 1973) possibly less available

food. It would be interesting to investigate the hypo esis
urban populations have smaller home ranges than t eir ,

counterparts, which appears to be the case even
Human activity and construction may provide or a ,’S ,

of food and cover resources and artifacts in t e
might be used by animals on occasion-all of which would faali

"" rh^h^trio kept to a fairly regular H-

tended to stay within a predictable home ran^ ^hid never been

they twice went into two separate areas ‘ jniolcrant

seen to go before. In one of these, resident ^ogs

toward the trio and actually chased them ranee- these

contrast to the dogs that lived within the temis with the

dogs being more tolerant, some even o hv huncer to

go into these new areas since they had r T ge

fact, forage for food in these new are ■ jub-

extending their effective home r^g ^

sequently return to these areas. The t ° noint after a night

cursion was 1.7 km (1.06 mi) measured point to point

of movement across their range.

52 -

THE dog: rrs domestication and behavior

MARKING BEHAVIOR

The frequencies of marking in various contexts are shown in Table I.
Both males marked more than twice as often as the female (F), who
never, in fact, marked in the presence of a strange dog. Marking in
the alleys of the home range and in the park were at a similar
frequency for X and F, but Y showed a greater frequency of mark-
ing in the park than in the alleys. The role of the marker may
change with a change in locale, X being the most frequent marker of
the home site area (alleys), while both males mark almost equally
in the park. Y marked more frequently than X in the presence of a
strange dog and was the most aggressive of the trio toward strang-
ers. The low incidence of scraping after marking in X may be due
to his hip injury; F was never seen to scrape. Marking over, that is
marking where another conspecific has just marked, may be a very
significant social phenomenon in the dog since it was one of the
most consistent findings in our study. Of the 36 times that F uri-
nated, one or occasionally both males were seen to mark over her
mark 30 times (see Table I). Twice, F marked over X. One occasion
served to demonstrate the intentionality behind this behavior:

4:15 a.m. X crosses the road while F pauses to sniff the

sidewalk. F urinates, then runs to join X but X sees
her and comes back to mark over. He then crosses
the road again and continues on his way with F and
Y following.

X marked over F three times as often as Y; this may indicate a
closer allegiance with the female F. Interestingly, X marked three
times over Y after Y had marked over F. Y was only seen to do this
once (marking over X's mark over F). What stimuli initiated mark-
ing in the first place were not determined. Often a horizontal sur-
face was marked (grass clumps, lawn, edge of pavement), but
more often a vertical object was marked (wall, comer of a wall,
tree, fire hydrant, lamp post or trash can). Marking was particu-
larly evident when the trio entered the park early in the morning.

LEADERSHIP

The observed frequency of which individual led the trio when mov-
ing within the home range after resting or eating is shown in Table

Behavior and Ecology of an Urban Feral Dog

53

*ms. m.uVs and ^cr.ipvs

54

THE dog: its domestication and behavior

Table II.

FREQUENCY OF LEADING GROUP

After resting, eating, and

Animal

ranging in park

Leads chase in park

Y

1

15

X

12

1

F

37

13

II. The frequency of leading while ranging in the park and while
chasing squirrels is also included. Clearly, in the home range, the
female (F), was the usual leader. X usually followed F, and Y would
then often follow the pair, possibly a magnet effect (see later). Male
X tended to lead to the lake in the park where the trio would drink
and swim and was also the one to lead the group back from the
park (see Figure 5). Male Y tended to lead active chases in the park.

Leadership entailed a number of subtle overt signals, espe-
cially eye contact. For example, F or Y would wait for X to catch up,
often sitting and looking back at him. As soon as they looked back,
he would speed up. After resting, F would stand up, walk off a few
paces and then turn and look back at X and Y. This looking back
stimulated them to follow. F would then run or trot a few more
paces and look back. If they were not yet following, she would
stop, look back and then give an exaggerated head turn in the
direction she was heading. The following anecdotes from our field
notes will further clarify this subtle communication:

2:50 a.m. Trio sleeping on porch. F gets up, gives a yelp bark
and goes down steps onto sidewalk with Y. X sleeps
on. F returns 5 min later and gives another yelp bark
and a tail wag toward X. X gets up and follows her.

F face-rubs and tail-wags with X.

3:55 a.m. Trio sleeping. F stands up, looks at Y. Y im-
mediately sits up. F gets off paHo and goes onto
sidewalk and looks back at X and Y. She then sets
off down the road. Y stands up. F stops and looks
back at Y. Y now follows her. X still sleeps. Y looks
back at X, then follows F. After 5 min, Y and F re-

Behavior and Ecology of an Urban Feral Dog

55

turn to patio. F tail-wags at X who is still sleeping. Y
sits close by and tail-wags. X sits up. Y moves off. F
waits by X. Y returns to porch and tail-wags at F. F
goes onto sidewalk and looks back at X. X stands
up, marks twice and Y marks and scrapes. Y crosses
road and looks back at F. F (in middle of the road)
looks back at X, who joins her "at last." She gives a
tail-wag, face-bump and exaggerated head turn to X
and then leads him across the road to join up with
Y. With F in the lead, Y and X follow her to forage.

The following interacHon illustrates further the dynamics of
communication between two members of the trio. X is wit a resi
dent free-roaming female who is in heat.

8:30 a.m. F moves off and looks back at X. X looks at

lows her for a few paces, then stops and looks back
at the estrous female. He then returns
female. F moves off again a ® * F

and looks back at X. X looks at F and then follows F

back to the flaphouse.

The magnet effect of the trio keeping together is exemplified
by these two observations:

4:30 a.m. F crosses the main highway but X -d Y stay on the
sidewalk and do not follow; F returns to them
all three go off in a different direction.

7:15 a.m. Y crosses the main and d^not fol-

park; X and F stay on the other two.

low. Y looks back and then rejoins the other tw

Further notations on on^hunHng acHvity in'^the

sented subsequently in the observatio
park.

MISCELLANEOUS SOCIAL INTERACTIONS

eve contact (looking

In this category, overt and waiting for a con-

at), and covert communication of to i„.dership). Only overt

specific, are not included (see section o

Figure 5. Trio behavior in the p
bathing and drinking in lake, (d) i
one spatial pattern while looking
scraps, and (g)F and Y wailing fot
before leflviwp park.

DOMESTICATION AND BEHAVIOR

sodal behaviors associated vv'ith ^ considered. Among

aggression, and social investigation wiU ® recorded

members of the trio, a total of only i6 in .chance of this un-
tliroughout the entire study. The possi ® ^ discussed sub-
expectedly low frequency of overt be

sequently. observed behveen X, F,

No displays of submission were e erovvied at X and Y,

and Y. On two occasions during feeding, interactions,

respectively, and both withdrew. Of the 14 included bnef

all were friendly and of extremely short dura nose-

taU-w;,a.Hr,cr face-bumping or ruboing.

were friendly and of extremely sno tnhbine

1-wagging, play-bow, face-bumping or

58

THE dog: its domestication and behavior

DUshine toward the mouth of a companion. One bnef (5 sec) play-
*ase 4s recorded in the park between F and ^

summarized in Table 111. X initiated no mterachons himself engag-
ine in one redprocal face-rub and tail-wag with Y when the latter
rejoined the trio after a brief absence. Possibly an indicator of soaal
preferences, the female F initiated more interactions toward X than
toward Y, while Y interacted almost equally with X and t.

Interactions with other dogs, some identified as free-roammg
residents that shared the same home range as the trio and
nonresident free-roaming and feral dogs were as follows: Ot 33
recorded encounters, sb< were designated neutral where the
stranger approached the trio or passed by and was ignored. Eight
interactions were judged as friendly, since reciprocal tail-waging
and, more rarely, play acHons were observed (see Table IV). Nine-
teen other interactions were aggressive or offensive. On one of
these occasions, a dog crossing into the park was intimidated by
the mere sight of the trio and ran off (none of the trio even looked
at the dog), and it was hit by a car. More often the strange dog would
be intimidated by Y and would run off. On eight of the aggressive
occasions, Y was the main instigator of intimidation toward strangers
(see Table IV), On one occasion F supported Y in chasing a stranger.

Table HI.

FREQUENCY OF INTERACTIONS WITHIN THE TRIO

Recipient

Initiator

X

F

Y

X

-

-

RF“

F

A**

F F F F F* (5)

-

A

RF

y

F F F F (4)

RF

F F F (3)

RF

-

•RF, r«?ciprt>cal friendly interactions.
'A, agonistic interactions.

*F, friendly Interactions.

Table IV.

SUMMARY OF VARIOUS INTERACTIONS WITH STRANGE DOGS (Sd)

Threats

Y barks at SD 6 and chases;

F supports Y, barks and chases.

Y then urinates.

SD approaches and stands still. Y
alerts and looks; SD ? runs off.

F barks at SD. Y chases SD and
then returns and tail-wags to

X.

Mange-covered SD 6 challenged
by F. No submission, briefly
follows trio. Ignored by X and

Y.

*SD <5 approaches, chased by Y.

Y approaches SD 6 who runs off;

Y marks.

SD 9 approaches submissively;
X, supported by F, chases her
away.

Y chases approaching SD ?, then
marks.

SD ? sees trio, retreats though no
obvious threat. Y marks.

•SD 6 approaches; X tail up, side
on threat. Y play-bows to SD,
then rushes and bites scruff.

SD runs off. Y play-bows to X,
then marks.

Friendly

Y approaches resident 6, recip-
rocal tail-wag and genital
sniffing

Y approaches resident d, genital
sniffing, then Y marks.

•Y approaches SD 7, investi-
gates, then gives play bow and
F prances.

Y approaches with tail-wagging
and investigates SD ?. F ap-
proaches, X follows and SD
runs off. Y marks.

•F meets 2 6 SDs; social investi-
gation and reciprocal tail-wag
but F intimidated until Y joins
her and greets the 2 ds. X
indifferent.

SD 9 approaches, gives low tail-
wag to Y, who reciprocates
with high tail-wag. SD 9 also
tail-wags to F and X. Trio leave.

•Trio meet SD d, SD 9 and SD d
pup; reciprocal greeting and
social investigation. SD 6 pup
joins trio briefly.

■Symbols: ?, indicates sex of SD not determined; •, occurred in park. Ail other
observations in alleys.

58

THE dog; rrs domestication and behavior

pushing towatd the mouth of a companion. One brief (5 sec) play-
chase was recorded in the park between F and Y. These data are
summarized in Table III. X initiated no interactions himself, engag-
ing in one reciprocal face-rub and tail-wag with Y when the latter
rejoined the trio after a brief absence. Possibly an indicator of social
preferences, the female F initiated more interactions toward X than
toward Y, while Y interacted almost equally with X and F.

Interactions with other dogs, some identified as free-roaming
residents that shared the same home range as the trio and others as
nonresident free-roaming and feral dogs were as follows; Of 33
recorded encounters, six were designated neutral where the
stranger approached the trio or passed by and was ignored. Eight
interactions were judged as friendly, since reciprocal tail-wagging
and, more rarely, play actions were observed (see Table IV). Nine-
teen other interactions were aggressive or offensive. On one of
these occasions, a dog crossing into the park was intimidated by
the mere sight of the trio and ran off (none of the trio even looked
at the dog), and it was hit by a car. More often the strange dog would
be intimidated by Y and would run off. On eight of the aggressive
occasions, Y was the main instigator of intimidation toward strangers
(seeTable IV). On one occasionF supported Y in chasing a stranger.

Table III.

FREQUENCY OF INTERACTIONS WITHIN THE TRIO

Iriifiator

Recipient

X

F

y

X

-

-

1 RF*

F

F F F F FM5)

-

A

RF

Y

FFFF(4) I
RF

F F F (3)

RF

-

*RF, reciprocal friendly interactions.
M. agonistic interactions.

'F, friendly interactions.

Behavior arid Ecology of an Urban Feral Dog

6i

fined resident dogs and the trio, while foraging in the back alleys of
houses.

On three occasions the trio howled to passing police car sirens.
They avoided people, especially children, Y having the greatest
flight distance. In the park, people were often ignored. Relation-
ships with cats were consistent. All three observed encounters re-
sulted in chasing, and on one occasion a cat was almost caught by
F. One rat kill was observed in a back alley; this was the only live
prey the trio was ever seen to secure.

PACK COHESION (FISSION AND FUSION)

The frequencies that any individual was not in the trio for any
period of time greater than 3 min were as follows. F broke the trio
twice for 5 min, once with another nonresident male dog and once
to forage alone. X was not with the other two dogs on only two
occasions, once when he returned 5 min early from the park and
once when he refused to get up and continued to sleep for 5 min
before F and Y got him to follow them. Y was the most often away,
F and X being seen together in the absence of Y a total of 15 times,
indicating that Y may have had a lesser tie to the group. He would
often sleep alone in one of the houses. Since we do not know the
origins and earlier relationships of these dogs, no further conclu-
sions can be drawn. Usually no greeting or overt interaction was
evident when Y rejoined the other two dogs.

HUNTING BEHAVIOR

Shortly after the initiation of the study, the trio, instead of retiring
for the day around 6:30 a.m. would go to a park adjacent to their
home site for 1-2 hr. (Since they later stopped going to this park,
we may conclude that this was a summer schedule activity.)
In the park, the trio would scavenge profitably for food under
park benches and picnic tables, but most of the time in the park was
spent chasing squirrels. On entering the park and marking, they
would trot across open space and then break into a ranging run
beneath the trees. If any squirrels were seen on the ground, they

6o

I AND BEHAVIOR

Both Y and X were seen supporting F in intimidating an approach-
ing stranger.

More rarely, a strange dog would join the trio for varying
periods. One resident free-roaming dog, a male, was seen foraging
with the trio early in the morning on three occasions. Another
morning, a small male pup played with F in the park after meeting
the trio earlier in a "morning greeting" (described below) and also
briefly joined the trio to hunt squirrels in the park. The most in-
triguing temporary union of a stranger with the trio occurred one
morning. A male dog, about the same size as F, stayed with the trio
from 4:42 a.m.-7;40 a.m. while they foraged in the alleys. When
the trio returned to one of its houses, he remained outsi^._The
field notes summarize what followed:

7:30 a.m. F comes to the front door of the abandoned house
(Y and X have retired upstairs). The stranger, a
brown male (Br), looks up at F, tail high and wag-
ging, and marks the porch. F comes down the steps
and leads Br to the flaphouse. He follows her up
the steps. Meanwhile X comes out of the back door
of the first house and trails the pair. X and F go into
the flaphouse while Br waits on the steps.

His approach/withdrawal behavior in front of the door indicates his
ambivalence about entering the unfamiliar house. After waiting 2
min, he leaves. This dog was never seen again by the observers.

Toward the end of the study, a young female dog joined the
trio for an estimated 6 days; it was subordinate to F but was ac-
cepted by her.

The trio was seen once involved in one of three observed
morning greeting ceremonies where several dogs come together
and engage in reciprocal social investigation and greeting (see
later).

The trio was on friendly terms with two separate dogs who
were usually tethered in their backyards. During the early hours,
the trio would enter these dogs' backyards, were usually greeted,
and would eat out of the dogs' food bowls and forage in the gar-
bage cans in the yards. They would often rest or rendezvous after a
brief separation in these yards and use one as a regular crossing
betrvoen streets. Tail-wagging was often noted between other con-

Behavior and Ecology of an Urban Feral Dog

63

The dogs expended considerable energy on these mornings in the
park. The adaptive significance and bioenergetics of this hunting
behavior are questionable since out of a total of 61 chases (see Table
V) no squirrel was ever caught. This activity may be reinforcing
itself, i.e., like play.

OTHER DOG GROUPS

The occurrence of dogs seen alone and in groups of two or more in
the area of study and in adjacent regions is summarized in Table
VI. Most dogs were seen singly, although pairs of dogs were com-
mon. It was not possible to sex all pairs observed, but where it was
possible, either two males or a male and female were together. No
female pairs were identified. SN trios of dogs were seen, two of

Table VI.

OTHER DOGS OBSERVED IN THE STUDY AREA

Number of dogs together

3 2 , 3 4 5 6 7

Resting,

moving or foraging 161 30 6

With $ in heat _ ~ 4 groups

of 3 <Js

9 with pup(s) — with —

1 pup

Pups alone

without mother — 1 ^

With people 14 —

“Greeting" group — —

(briefly
foraging in
same vicinity)

— 12 4

9 with — — —

3 pups (X 2)

Total count

175 64

33

8

10 18 35

62

THE dog: its domestication and behavior

would immediately give chase. The trio would systematically work
over different areas of the park where there were stands of trees.
Once, and often twice each morning, they would drink and bathe in
a large lake in the center of the park for 2-3 min. Bouts of chasing
and ranging were interspersed with resting periods, when the trio
would lie or sit in three patterns of head orientation (see Figure 5).
These patterns optimized the chances of spotting a squirrel and as
soon as one dog alerted and oriented (and occasionally tail-
wagged), so would the other two. If it then gave chase, the other
two would follow immediately, correcting their orientation when
they too saw the prey. This later orientation often resulted in one
dog bumping into a companion or momentarily running in the
wrong direction. Also, while in a ranging run if one dog suddenly
ran to one side upon seeing a squirrel, the others would follow at
once and then look. If, for example, the center dog in Figure 5e
suddenly alerted forward, the dogs on each side would alert in the
same direction. If the dog on the far right in Figure ye alerted in a
forward direction and to its right, the other two dogs would turn
around quickly and orient to their left.

Chases were usually terminated by the dogs jumping up at the
tree in which the prey had sought refuge (see Figure yd). F and,
less often, Y would jump up against the tree, and barking and
taU-wagging were frequently noted. On two occasions, squirrels
were seen to jump out of the trees up which they had been chased,
while the dogs were barking and jumping up at the base of the tree.

Table V.

PARK CHASES

Date

Tanperalure

Time of onset

Number of chasesIHme .

June 8

70°?

5;15 a.m.

6 chases/190 min

June 11

76°?

5:35 a.m.

30 chases/170 min

June 13

73°F

(Rainy)

5:33 a.m.

10 chases/180 min

June 15

73T

5:15 a.m.

10 chases/225 min

July 9

77°F

5-35 a.m.

5 chases/120 min.

Behavior and Ecology of an Urban Feral Dog

Discussion

65

Against the background of Beck's (1971, 1973) studies of urban
dogs, the present study focuses more specifically on the social
ecology of a pack, or social unit, of three animals. Permanence of
such a social unit may be rare in urban dogs, but this conclusion
awaits further study. The distribution and abundance of food (i.e.,
wide dispersal, but in small concentrations) would favor solitary
scavenging, and all other feral dogs observed during this study were
seen singly and more rarely in pairs.* Groups of three or more were
usually temporary packs of males following a bitch in heat. A per-
manent pack in the urban environment would also be conspicuous
and more liable to human predation (the local dog catcher chases
groups more than individuals).

In relation to the diurnal activity of other dogs, the feral trio
remained undercover during the major part of the day. This may
have been to avoid the heat of the day or human interactions, or
both. They clearly avoided close proximity with people, while the
activity of nonferal dogs closely coincided with human activity,
i.e., a morning peak around 8:00 a.m. (when many are presumably
let out of the house) and a peak in the early evening when dogs,
people, and children would be out in the streets. Nonferal dogs
were least active during the afternoons, the hottest time of the day,
and, like the feral trio, were most active during the early hours of
the morning.

In contrast to the regular morning visits of the trio to the park
to scavenge for food and to chase squirrels, other dogs were only
occasionally seen there and without any regularity in their visits.
The amount of time and energy the trio spent in ranging over the
park and chasing squirrels, which they were never seen to catch,
was remarkable. Such unreinforced activity may be rewarding in
its own right and could be interpreted as play.

The interacrions between the three dogs were surprising, since
much more overt communication and social interaction were an-
ticipated. There was little "redundancy" however, and much

•Feral and free-roaming dogs observed in Mexico Cily and Madras n ere seen forag-
ing singly and only occasionally in pairs (M W Fox)

64

THE dog: its domestication and behavior

which were foraging, two playing, and two traveling. Three early
morning greeting groups of respectively 5, 6, and 7 individuals
were seen. These groupings were of extremely short duraHon (1-3
min). Groupings of longer duration were focused around a female
in heat: four groups of 3 males and 1 female, one group with 5
males and 1 female, two groups with 6 males and 1 female and
three groups with 7 males around 1 female. A major stimulus for
temporary pack formation in the urban dog is clearly the presence
of an estrous female (see Figure 6).

Figure 6. Free-warning house dogs forming a temporary pack
awund an esiwus bitch (a), creating a traffic hazard crossing
highieay (h>, and attempting to reach the bitch who is seeking ref-
uge under an automolrile (d). The resident was unable to leave the
house for fear of being attacked as the temporary pack rests on her
front steps (c).

Behavior and Ecology of an Urban Feral Dog

67

disease problems, both pre- and postnatally, would reduce the
chances of any offspring surviving up to weaning age. Toward the
end of this study, F was reportedly in heat (September, 1973) ac-
cording to one local resident, and further observations confirmed
that she was pregnant (or pseudopregnant). If any pups were
bom, they could not have survived long since we were unable to
find them although the trio was still active in the area through
January. While Nesbitt (1975) has shown that there is some natural
selection in feral rural dogs for a particular body size/phenotype, it
is unlikely that natural selection could operate in feral urban dogs
because of the high turnover rate of the population and low chance
of survival for any offspring.

Summary

A group of three feral dogs (two males, one female) living in vacant
buildings in St. Louis, Missouri was studied. They avoided close
proximity with people and were active earlier and later in the day
than the people and loose pets in the area. They found food while
scavenging throu^ human trash. The group'^s activities were usu-
ally initiated by the female of the group though otherwise there
was no clear linear hierarchy and few ritualized displays of domi-
nance or greeting. Specific roles within the group were observable.
Though the female appeared to be pregnant during the study,
puppies were never noted.

66

THE DOG: ITS DOMESTICATION AND BEHAVIOR

communication was overt, involving body orientation and eye con-
tact. It must be concluded that each member of the social unit was
fully aware of what the other dogs were doing and were intending
to do much of the time, i.e., a metacommunicative system and
context-related set of expectations were operating. In addition,
since there was no clear linear dominance hierarchy, there were
few ritualized displays of dominance and submission and no greet-
ing rituals to the alpha individual (as exemplified by the wolf pack,
[Fox, 1971b]). Leadership, i.e., choice of direction of movement,
was usually made by the female, who at such times could be desig-
nated leader. On two occasions (while feeding) she asserted domi-
nance over the others by staring and growling. The response was
nonsubmissive avoidance. Competitive behavior was never seen
on other occasions.

In relation to context, therefore, individual differences were
evident, and in different contexts the female was the leader or was
dominant. If behavior differs between individual members of a
social group in the same context, then the possibility of context-
related roles in relation to the social ecology of the group must be
considered. Two of the dogs of this social group appeared to have
different roles which were apparent at different times in certain
contexts. While the female was most often identifiable as the
leader or decision-maker, the male (Y) was the most aggressive
toward strange dogs and might be designated as the guard of the
unit, A specific role for the other male (X) was not so easily as-
signed. There was no evidence of dominance over, or subordina-
tion to, the other male, and X marked (the home range) 2.6 times as-
frequently as Y (except in the park) and marked three times as
much as Y over F after she had urinated. This latter observation
may indicate a closer bond behveen X and F than between F and Y,
an interpretation supported further by the fact that Y left the social
unit more often than either of the other dogs. Also the female (F)
initiated more social interactions with X than with Y, while Y in-
teracted at almost equal frequencies with X and F. There was no
ewdence that the marks were cither used or respected as territorial
boundaries in this feral trio.

In spile of these feral dogs' remarkable abilities to adapt to the
urban environment, it is unlikely that they could succeed in raising
a litter of pups under such conditions. Nutritional, parasite, and

IV

Vocalizations in Wild
Canids and Possible
Effects of
Domestication

Introduction

In order to make an inventory of the various sounds emitted by
canids, the vocalizations from several different species of Canidae
reported in this study were recorded, and selected sounds were
analyzed on a sound spectrograph. From such an inventory, m-
terspecies comparisons of vocalizations and of the vanous contexts
in which they occur could lead to a closer understanding of
taxonomic interrelationships. It might also throw I^ht on the on-
gin(s) of the domesticated dog Cams fmmlmns which is, to dale, an

^™^“e study of canid vocaUzations is also of significance for other

reasons. Firsh an understanding of *”''ntcT

mals is a necessary part of understanding their total communica-
tion system (Fox and Cohen, 1977). foj 'he
enhance visual and olfactory displays but in

as a substitute for these displays, as f domic

cover, or at night. Second, the canid family is

in some respects. Its members come from a " 'do v. ncU) o sr^al
systems (Fox, 1975) ranging from the greganous (eg., ivohcs.

Vocalizations in Wild Canids

71

Table I. continued

Yowl-howl

Coo

Growl

Cough

Bark

Click

Toothsnap

Pant

WD

F

WD

—

D

—

—

F

—

—

—

—

D

—

WD

FD

—

—

—

—

—

—

—

—

W

_

WCDF

WCFD

WD

F

WCD

—

—

F

WCDF

WCFD

WCDF

D

F

WCD

—

—

D

—

D

—

—

—

WCD

F

—

—

D

-

—

-

WCD

—

WCD

—

WCD

—

-

-

•Most sounds are not mixed, but some of the sounds may be mixed simultaneously
or successively, e,g., whine-growl, bark-howl, yelp-bark.

‘F, red, arctic, and grey fox; D, dog; W, wolf; C coyote; N, neonates of these species.
•Also "contentment grunts" of contact in neonate W, C, and D.

•“May serve as warning to others.

3 wolves {Canis lupus) from birth to 2 years

2 red foxes (Viilpes vulpes) from birth to lo weeks

4 red foxes from 2 weeks to 10 weeks
4 red foxes from 2 weeks to 3 weeks

2 grey foxes (Urocyon cineraoargentcus) from birth lo 10 weeks
2 grey foxes from 8 weeks to 2 years
8 Chihuahuas from birth to 6 months

4lrishsetterxDobermanpinscherF.hybndsfromb,rthtowecks

4 coyotes (Cams latrans) from birth to 3 years
4 F. coyote X beagle hybrids from birth to 3 years
2 ArcHc foxes (Alopex lagopus) from 8 rveeks vmrs

2 Asiatic golden jackals (C. aurens) from i week to - > cars

70

THE dog: its domestication and behavior

Table I.

CANID SOUND TYPES®

Social Context

Mew

Grunt

Whine

(whimpers)

Yelp

Scream

Yip

Greeting

pb

WCD

D

F

C

Play-solidt

—

—

D

D

—

—

Submission

F

_

WCD

D

WCD

Defense

-

—

WCD

D

CFW

—

Threat'’

—

—

Care- or con-
tact-seeking

NF

DW

N;DWC

N of WCD; C

F

C

Distress (pain)

N

—

N; WCDF

N of WCD; D

N; WCDF

Contact-seeking
(lone calls)

N; WCD

D

Group vocali-
zation

—

WCD

c

rela«vely nonsocial (e.g., foxes). Com-
fFnif "'m greatly between these two main groups

box, i9^b), and comparative studies may add to our knowledge of

knowSlHoo ^ evoIuHon of canids. This.

icement^fT^J essential to the preservation and man-

agement of endangered species.

Materials and
Methods

Over a penod of 7 years, recordings of vocalizations were made of
the following speaes at various ages (the animals were all hand-
raised m captuTly):

Vocalizations in Wild Canids

73

from more complex mixed sounds, which are described shortly. It
should also be emphasized that under the categories of whines and
coos, for example, there are clear differences between and within
species in the pitch, duration, and spectrographic structure of such
sounds. The focus of this study, however, is not to look at such
species and individual differences within sound types but rather to
make an overall survey of the occurrence of these basic sound
types across species to provide a basis for evaluating the possible
effects of domestication on canid vocalizations.*

The cross-species survey clearly reveals the distinction be-
tween vulpine (foxlike) and canine (dog/wolf-like) species, the
former showing a high incidence of coo-calls and guttural clicks
(absent in the latter), a high frequency of screams, and an absence
of group vocalizations. The yip sound type was exclusive to
coyotes and golden jackals, and a high incidence of barks in many
different contexts was characteristic of the domestic dog. No var-
iations in occurrence of sound types in individuals of the same
species were ever recorded in different contexts, i.e., sound type
and context specificity was evident for each species studied.

In addition to these vocal sound types, aU species commurii-
cate to some extent by means of mechanical or unvoiced sounds
such as the tooth-snapping heard in wolves, coyotes, and dogs,
and panting, a play-soliciting signal of dogs and foxes (see page
135). Because these mechanical sounds are of relaHvely low volume
and do not carry very far, they are reserved for close contact situa-
tions and are often heard during agonistic encounters.

CROSS-SPECIES COMPARISONS

Some evidence is available at this point to warrant preliminary
cross-species comparisons (see Figures 1 an z).

t differences, which is in itself

"This is not to underrate the significance o ‘ ^lear indi^Tdual dif-

an important and much needed area as for example, wolves engaged in

ferences became evident during the s j,fferenl sexes emitting the vime c«x>
group howling and in red and grey oxes duration Such vanations could

rail in nifrh. form, intensity, ^ state and intentions

call varying widely in pilch, fhe emotion-i

permit even more information excha g
per se that are discussed in this chapter.

7 ^

the dog: its domestication and behavior

Recordings were made in nine different social contexts listed in
Table I. In addition, recordings were also made of various zoo
animals including Dingo {Canis exfamiliaris dingo). New Guinea
singing dog (Canis exfamiliaris hallstromi), maned wolf (Chrysocyon
brachyitrus), culpeo (Dusicyon culpaeus), wolf (Canis lupus), coyote
(Canis latrans), bush dog (Speolhos venaticus), and Cape hunting dog
(Lycaan pictus).

All recordings were made on a Sony EM-2 field recorder at 7
in. per second. South analysis was then performed with a Kay
Sona-graph 6061A, with high shape (HS) and narrow band set-
tings.

Later references to short, long, or extended sounds refer to
arbitrary classes of sound duration as follows; short, 0-0.9 loTig,
1.0-2.4 sec, extended, 2.4+ sec.

Results

CLASSIFICATION OF SOUNDS

Twelve basic canid sound types were identified in this study includ-
ing whines (and longer softer whimpers), short yelps, yips,
screams, barks, coughs, growls, coos, howls (or yowls), mews,
grunts, and guttural clicks. Variation within these categories may
occur at both the individual and interspecies levels, and sounds
may vary along the dimensions of duration, frequency, intensity,
cycle (rhythmicity), and context.

Not all of these basic sound types arc included in the vocal
repertoires of every canid species, and the same sound may be
used by different species in very different contexts. Foxes, for
example, are the only canids known to emit a pure scream in greet-
ing conspecifics, and domestic dogs will bark in many situations
(c.g., threat, play-solicit, contact-seeking) while foxes bark only in
threat. The presence or absence of these various sounds in dif-
ferent canid species is shown inTablel together with the context(s)
in which each sound was recorded. Further descriptions of some of
these basic sound types may be found in Tembrock (1968, i960),
Bleichcr (1963), Joslin (1966) and Thebergc and Falls (1967).

The basic sound types listed in Table 1 are to be differentiated

Figure 2. Mixed sounds. Successive: (A) bark-^growl of a 4-
week-old male Chihuahua: (B) pure coo and coo^scream of a
i^-day old female red fox; (C) yelp-rgroiol-rbark-^grou>l of a
lo-day-old male Irish setter X Doberman pinscher. Stmultane
ous: (D) bark-howls of a i-year-old female wolf: (E) growl-
scream of Culpeo: (F) growl-scream of a y-month-old female
Arctic fox.

cated a fundamental and PDF of 1500-3400 Hz. ^hirang is most
often a cyclic (i.e., rhythmic) vocalization given m ^stress, with
the exception of the undulaHng whine which is noncyclic.

"'"'^hese sounds, as shown by Bleicher (^963). ^lop in
in combination with the whine ^vhine ^discussed

successively combined with a grow , ' in other adult

later). This particular sound type ^ ^ ■ familiaris. For

canids and may be a species-charactenstic 01 cai. 1 hortened,

the purpose of this study, the yelp d piercing variant,

contracted form of the whine; a high jackals.

the yip, being an ®°"^/g“[ext in the fox is the scream.

The analogy of the yelp in the same c

74

THE DOG; ITS DOMESTICATION AND BEHAVIOR

Figure 1. (A) Whines of a , -day-old male Chihuahua. (B)
Screams of a 6 -day-old male Chihuahua. (C) Barks of an adult
fmale dingo. (D) Crawls of an adult male wolf. (E) Coos ofan-
ioy-fd female red fox. (E) Excerpt from a howl ofan aduti female
wolf (C) Mrros of e y-week-old male Chihuahua. (H) Grunts of
an 8 -day-old male Irish setter X Doberman pinscher hybrid.

ratiol^ I i !f short du-

Hmes “bservaKon) will some-

hmes extend this sound for several seconds, however orodudne

Tound whine (Crisler, '1958) Thif

cadw altemaM W w ‘ongue within the vocal

ravity alternately blocking and opening air passages (Crisler, 1958).

the whiner'^f <* 0 "''nant frequency (PDF) found in

the whines of adult wolves were both about 1570 Hz TTie same

figures apply to the undulating whine nz. me same

whill^he PnP™""'"' u ^ ^"Ses from 400-1570 Hz,

uhile the PDF is usually from 2000-3000 Hz. Recordings made of

developing Insh setter x Doberman pinscher hybrid puppies indi-

Vocalizations in Wild Canids

77

but not analyzed, in all canids. It may serve as a warning to off-
spring and others in agonistic contexts.

Howls

All recorded howls of wolves and coyotes were of long to ex-
tended duration. The fundamental for both species was between
400-2000 Hz and the PDF from 1200-2900 Hz.

Mews

Preliminary evidence indicates, that newborn to 5-week-oId
red and grey foxes tend to repeat their neonatal mews more often
than dogs of the same age group. Foxes maintain this sound as part
of their vocal repertoires throughout life, while it is only heard in
neonates of all other canids studied.

The fundamental of the grunts of wolves and dogs ranged
from 85-200 Hz when detectable. The sounds are heard in neon-
ates of these two species and the coyote but were not recorded in
foxes.

This sound is heard only in foxes. One fom is trill 1 ^^
the other is more of a cackle. Differentiation e
on the basis of spectrographic evidence is difficult, and it is not yet
clear which stimuli elicit each form of the coo.

MIXED SOUNDS

The above sound types may be mixed

superimposition of two vocal sounds o y , ; n /-pp Fieure

tw"o or m"ore types of sounds (vocal andf^rmech^

2). A combination of these two mixe follow a pure

a mixed sound, for example, ""'S correlafes ivell with the

sound. The phenomenon of sound m g expressions

superimposiKon of body 'artgoage pos eliciting

(Fox, 1971b). A wolf faced with a situahon s

•Tliis sound type is to be distinguished from the gntuj end gntn.s of u s.cl, nr dung

canid, or of one in intense pain as dunng pa

76

THE dog: its domestication and behavior

Screams

The screams of red foxes tend to be of longer duration and
occur in a greater variety of contexts than those of other canids.
The gray foxes recorded emitted screams which were generally
shorter than those of the red foxes but were repeated more often. It
is thus possible that the total signal value of the scream for these
two species is about equal. The fundamental for most foxes studied
(including the Arctic fox) ranged from 1200-2000 Hz. The PDF was
between 2000-5000 Hz. Chihuahua recordings indicate a funda-
mental of 1200-2000 Hz and a PDF of 2000-3200 Hz. Recordings of
Irish setter X Doberman pinschers show fundamental and PDF val-
ues between 1800-2700 Hz. Preliminary data on the screams of
coyotes indicate a fundamental of about 2400 Hz and a PDF of
about 2700 Hz.

The scream is noncyclic in red foxes and repeated, but not
cyclically, in gray foxes. Chihuahua puppies given painful skin
stimulation screamed noncyclically. Those placed on a cold sur-
face, however, emitted cyclic screams.

Barks

Barks of all species recorded were of very short duration (i.e.,
0.5 sec). All principal frequencies (fundamental and PDF) lie in the
lower register between 0-2000 Hz. The main differences between
the barks of various canid species concern cyclicity. The domestic
dog will often bark cyclically in a singsong manner, one bark fol-
lowing another until a train of barks results. This is commonly
heard during territorial defense and care or contact solicitation.
Foxes bark noncylically, while our data indicate that wolves may
bark either cyclically or noncyclically. Further investigation is
needed to distinguish the different stimulus situations that elicit
these two barking forms in wolves.

Growls

Growls may vary in duration from short to extended, depend-
ing on the situation and intensity of the social encounter. The
growls of all species are noncyclic. While foxes growl only in threat
and defense, wolves and some dogs growl while greeting one
another possibly reaffirming their dominance relationships.
Wolves, dogs, and especially coyotes may often growl during
group vocalizations. A muffled growl-bark or cough was recorded.

Table III.

MIXED SOUNDS BY SIMULTANEOUS EMISSION OF TWO SOUND TYPES

Vocalizations in Wild Canids

79

I I I

u 1 I

.. I I

I

5

I I
I I

u

o

S

I I I

O Q

I I I

n.

I

I I
I I

I I

I I

i d C

.ibbrcvl.Mcd form~lhc yelp and yip.

*D. dt^meMte doR, W. woU: C. coyole; F, red fox.

78

THE dog: its domestication and behavior

1 I 1

o

z

3

. o
’ci o

^ P4
^ K

s

(3

e

u

5

a

I

o

I I I

5

u

Q

1

u

2 :

o

u

Q O

I I

^ Q

u

Q

§ -2 I

<>i « o

1 I

I: I I

O a: 5 u

uoi9$!ua pimos ;sjy

I

lb

I I

I I I

1 1

li.

1 I I

I I I

I I

I I I

I I

¥ s? -s

•Complex combinations of more than two sounds are also possible, e.g., cough -*
click -* growl-scream

‘’Includes abbreviated form — the yelp and yip.

'D, domestic dog; w, ,wolf; c, coyote; f, red fox.

•‘Howl preceded by whining in wolf and yipping in coyote but not strictly mixed.

Vocalizations in Wild Canids

8i

Figure 3. (A) Complex long grunt of a i -day-old female
Chihuahua. (B) Yelp of a lo-day-old male Chihuahua (O Tran-
sitional bark-yelp of a 2-week-old male Chihuahua. (D) Bark of a
5-week-old male Chihuahua. <E) Sharp yelp with some vertical
Stratification (final segment) of a i-day-old male Ins se
Doberman pinscher. (F) Whine with coo component of a 26-day-ol
male Irish setter x Doberman pinscher.

ce of these souirds at birth, survival of the pups could be err-

" Mfday of age, the Chihuahua may a sound^wWch^is

St described as an extended grunt. , a Wt is wide-

e sound is more complex than the grunt ( ^ „ when the

nded with several frequency variaHons and occurs when

imal is handled. . . j freaucncv

By day 6 the pup's vocal capacty
nations increase. At day lo sharp freque

undingmost Uke a yelp. approximates those of the

m and frequency range of the nse PP , . may

lly developed bark. Thus, .t f,rrf,*k Furth^erLidcnce for this

a developmental precursor of the bark, rur,

mes from an intermediate stage weeks of

hich can be heard (or better, seen spcctrograph.cnll) 1 at .

8o

THE dog: its domestication and behavior

fear and aggression may concurrently display facial and body
characteristics expressive of both motivations (Schenkel, 1947). The
accompanying vocalizations are likely to include components that
may be heard separately in pure form in fearful and aggressive
contexts.

Tables II and III indicate the mixed sounds that have been posi-
tively identified, and it is likely that further investigation will reveal
more sound combinations.* (See also Figure 2.)

Less social canids such as foxes do not mix sounds to as great
an extent as do the more gregarious canids (see Table II and IH).
Foxes tend to live solitary lives and coming together briefly during
the mating season; thus much of their vocal communication relies
on calls that must be audible over long distances in order to be
effective. Since these animals spend relatively little time in social
units, a refined, graded system of low intensity vocalizations
would have low selection priority. Gregarious canids such as the
wolf and dog, however, tend to mix sounds more than the solitary
canids, and facial and body expressions are also more highly dif-
ferentiated (Fox, 1975). Thus, with increasing social complexity in
the Canidae, there is also an increasing complexity of the com-
munication repertoire.

SOUND DEVELOPMENT

We have recorded developmental sequences for two domestic dog
breeds: Chihuahua and Irish setter X Doberman pinscher. Data on
the sound development of the red fox are available in Tembrock
(1958)-

Oiihiialiim

The Chihuahua can whine, scream, grunt, and mew at birth.
All of these sounds are functionally similar as they serve to alert the
mother to the distress and location of her pups. Without the pres-

Vocalizations in Wild Canids

83

Discussion

The 12 sound types presented here represent an attempt to classify
the major components of a complex vocal communication system
in Canidae.

Developmental data indicate that the earliest sounds present
in canids are those which elicit approach of conspecifics, specifi-
cally the mother. These sounds serve to decrease social distance.
Once this distance is decreased, the mother may become aware of
the needs of her young and may then serve these needs. On sev-
eral occasions while recording the sounds of neonate domestic
dogs, the mother was visually isolated from her pups but well
within hearing range. As soon as the pups began to vocalize their
distress to various experimental conditions, the mother made
vigorous attempts to physically contact her young. Thus it seems
that the distress calls (e.g., whines, screams) are releasers for some
maternal behaviors, particularly retrieval. The w^ m w c e
releasing mechanism works is unclear, but it may be possi e a
the mother is (hormonally) predisposed (or sensitized) to tnfe^re
these sounds as noxious/aversive stimuli and acts m su a y
to cease them. „„

A similar interpretation may be applied to some
calizations. Captive coyotes have been observed emi ng g
screams when threatened at a distance by domman .

had previously attacked them. The defensive ° •

to cut off the attack. Here, again, the sound may be aversive or

noxious stimulus to the dominant animal.

The second set of sounds to develop are
withdrawal of conspecifics increasing soa

sounds (e.g., barks, growls, cUcks) are not^^^

develop between a and 4 weeks of age. A . _ pYolore their
becomfng more self-sufficient -d ^ ^nmng ‘o ejlomjh

tfte brradly dass^ cai^voca^s as ‘hose

82

THE dog: its domestication and behavior

age, when differentiation and mixing of sounds begins to take
place. By about the 4th week of age, the development of the bark is
complete (see Figure 3B, C, and D).

At day 18 a clear successive emission of a mixed bark-growl
may be heard. By the 4th week this combination of sounds be-
comes more refined.

Irish setter X Doberman pinscher

As in the Chihuahua, whines, screams, gruntsi and mews are
present at birth as distinct sound forms. At day 1 the sharp fre-
quency rise (yelp), appearing at day 10 in the Chihuahua, was
noted. In addition, some vertical stratification is visible from the
spectrograph (Figure 3E), indicating a possible relationship to the
growl. By day 5 frequency variations become greater. The first fully
developed bark was noted on day 10. Mixing also begins to occur
widely on this day as successive emission of sounds begins. Muting
by superimposition begins between the 2nd and 3rd week of age.
On day 26 a call was recorded when a pup was isolated. Spectro-
graphic analysis later identified the call as having the short vertical
frequency changes typical of the coo call of foxes. Cooing, how-
ever, is not heard in dogs in its pure form, and this recording may
provide further evidence for the evolutionary link between foxes
and domestic dogs (see Figure 3F).

Developmental data indicated that domestic dogs begin to mix
sounds first by successive sound emissions and later by superim-
position. The Chihuahua seems to develop its vocal repertoire
somewhat more slowly than the mixed breed of Irish setter X
Doberman pinscher. It should be noted that Doberman pinschers
have been specifically bred as guard dogs, and the sooner they
begin to vocalize the sooner they may be put to use. In addition,
the Chihuahua has been bred for neoteny, a state of somewhat
retarded development, and its schedule of vocal ontogeny may
thus have been affected to some degree by the domestication pro-
cess.

Vocalizations in Wild Canids

85

Successive bark-howls and short rolling howls comparable to a
barking stanza associated with defensive aggression in the wolf
have been recorded. This is interpreted not as an evolutionary
relationship between barking and howling but rather as a species
typical phenomenon, where howling in the wolf (like barking in
the domestic dog) may be at a low threshold and be evoked in a
variety of different contexts.

In addition to the A-eliciting and W-eliciting sound classes,
Tembrock (1968) suggests two further sound groups: warning
sounds and infantile sounds. The warning sounds (e.g., bark)

serve as an alarm call for the entire social group. Rather than in-
creasing social distance between members of the group (as with
W-type sounds), these sounds tend (potenhaUy or actually) to in-
crease the entire group's distance from an external danger. These
sounds often occur in the defense of group territoiy or when one
animal alerts its group members to a nearby predator.

The infantUe sounds are those which have become denved
and emancipated and reoccur in adulthood. The mewing sou no
adult foxes is one possible example of such a denve voca iz ,
as is the yelping of adult dogs (see later). The former yoca iza
occurs first in the neonate and serves to

giving behavior and later reappears during courts ip an ‘ ^

Figure 4 further iUustrates the relaHonship of ng ^d

W-eliciting sounds when viewed with respect to e ‘ intense
of moHvaHon. As moHvaHon for approach

A-type sounds may be emitted. Similarly^ as ^ emitted,

drawal increases, more intense W-type soun . ^ gpd

Certain sounds, however, cross over "e

may be heard in either A-eliciting "^' 77 * .' nn Screaming, for
sounds occur at the higher levels of mo ''“I ‘

example, may occur either as an mtense cojy a

threat vocalizahon m foxes. An an g sounds such

might be the observation that when hig jpp.irently m.ip-

as laughing or crying may cross over a

propriate contexts (Danvin, 1873). _,„HoMcal cons«iuences

The group howl in wolves can ha p. ■ lopi-ltu r

It is clearly an A-eliciting sound, 8 7 nd ac.oe

for its duration and being prece c .wrox lurib) Oncer-
submission-a coming together gre-etmg ntu.il (Fox. .07

84

THE dog: its domestication and behavior

A*eUciting

sounds

•Saeam

Coo-howl

Yelp-yip

Repealed

whines

Whines

Mew

Increasing intensity/arousal

W-eli citing
sounds

Growl

Cough

Bark

Repealed

barks

Barking stanza
'Scream

Hypo»ie//cii; relationships between approach- and
mIhdrawl-eUatmg sounds and increasing intensity tarousal
Paradoxically, screams, with intense arousal, may be noxious
(W-ehating) or atlenlion/care-soliciting (A-eliciting).

come evident (see Figure 4). Considering first the A-tvpes the

Ih^DDroa'^ch'^ T ‘ expression of a moHvation/need for

the approach of a conspecific. The more intense versions of this

HTt"scr:3r''1'in^ whine, "eirand con!

spoctrographically. ^ ^ features may be seen

inten^^ocMexnrn ^e the least

iy-se social

^\V-^vr»n^ + 1 • point ol the second conhnuum

•nd ddife™ b,„ki„g p„|p„|.,; ™„ a"

Vocalizations in Wild Canids

87

indicate that some type of social reinforcement or facilitation is
involved in achieving full vocalizing potential.

The fact that squirrel monkeys deafened at birth are capable of
producing all normal vocalizations indicates that the vocal reper-
toire is largely genetically predetermined. This seems to be the case
in domestic dogs as well since experiments by Fox (1971a) indicate
that no auditorially evoked potentials could be recorded from the
dog's brain until 2 weeks of age. In effect, the dog is deaf until this
age; but vocalizations begin to develop at (or possibly before) birth,
and the basic sound repertoire is complete by about 4 weeks of age.

In addition, hand-raised domesHc dogs fed and handled on a
strict schedule quickly decreased their frequencies of care-soliating
distress vocalizations, presumably because their sounds were in no
way positively reinforced by petting or feeding (Fox, i9^a).

When placed in unfamiliar surroundings, domestic dogs gen-
eraUy directed contact-seeking sounds toward the
often barking, whining, and yelping, while wld canids
be silent. This supports other evidence that there may be a geneh
predisposition toward dependency in the dog whic is no prese
in wild canids (Fox, 1971b). .c

Barking is another sound form which is of
it seems to have hypertrophied in the domestic og
used in a much greater variety of contexts than among w
As Table I indicates, dogs may bark dunng play

soUdtlng, threat, defense, _soficiting

and yelp-barks. This contextual vanety

itself may not always convey °™rific infoniiation to

tracts the attention of the receiver. The m P l ^_pJ5

follow would then be received through other

visual and/or olfactory. These cues ^vould then ident.f) the mean

ing of the accompanying barks. distinction

From the inventory of various sou . .u vuloine and

can be made between foxes The similarities

canine subgroups of the genus Cfliiis mg taxonomic

betxveen wolf, dog, and dine oHhc origin(s) of the

affinities but add little to our understanding 01 b

dog prior to domestication. The outstanding feature

86

THE dog: its domestication and behavior

tain occasions, however, the group howl may be aborted and ag-
gressive behavior may be released. This has been observed upon
numerous occasions in captive wolf packs and also among captive
groups of coyotes. A tape-recorded howl or a siren may be used to
evoke group howling, but at the point of coming together, any
conflicts present between group members may be reactivated as a
result of the ensuing close proximity.

A particular facial expression, open mouth "play face,"
coupled with panting in the domestic dog may be analogous to
laughter in man (Fox, 1971b) occurring in similar contexts, namely
greeting and play-soliciting. In the red fox, panting may be accom-
panied by muffled screams, mews, and purring sounds during
greeting.^ The excited panting, a metacommunicative play signal,
exeniplifies how emotional changes may influence the rate and
depth of respiration.

data, particularly that concerning mixed sounds, imply
that canid vocal communication is essentiaUy an emotional lan-
page, comparable to the intenHonality expressed in nonvocal
body postures and facial expressions. This is supported by the fact
gradaHons of sound types are possible indicating
he degree of arousal motivaHon and intentions in a given context

tonerof ‘o ‘he "paralanguage” or e^tional over-

l^mt moHvT " “"•espond to mixed or ambiv-

thafZT^ 1^"® f Experiments by Tembrock (1958) indicate
is more f =^P^™‘ed from their mother

throWKcT ‘he mother's retrieval behavior than

lendfsuDoo^ themselves. This, ton,

sdd thaTfs ‘‘ “ "^“ssarily so much what is

said that is of signihcance but rather how it is said

less frequently than those raised with conspecifics When intro-
duced at a later age to other animals, howeven these isXes beean

WiZT ^t'" f “f 1 ‘hose raised in sodal groups.

Winter el a! (1973) report the same phenomenon in fquinel
monkeys (Saimm ^„rc«s) which had been deafened at bWh or

of vSn^ r ‘hoh frequencies

of vocalizing when introduced to conspedfics. These observations

Vocalizations in Wild Canids

89

Developmental data indicate that domestic dogs first begin to
mix sounds by successive sound emissions at about 10 days of age
and later by superimposition between 2 and 3 weeks of age. The
frequency of occurrence of the basic sound types in different con-
texts varied between species but not within species. The possible
effects of domestication on canid vocalizations are discussed.

Vocalizations in Wild Canids

89

Developmental data indicate that domestic dogs first begin to
mix sounds by successive sound emissions at about 10 days of age
and later by superimposition between 2 and 3 weeks of age. The
frequency of occurrence of the basic sound types in different con-
texts varied between species but not within species. The possible
effects of domestication on canid vocalizations are discussed.

Vocalizations in Wild Canids

89

Developmental data indicate that domestic dogs first begin to
mix sounds by successive sound emissions at about 10 days of age
and later by superimposition between 2 and 3 weeks of age. The
frequency of occurrence of the basic sound types in different con-
texts varied between species but not within species. The possible
effects of domestication on canid vocalizations are discussed.

88

THE dog: its domestication and BEHAVIOl

tic dog — balking — ^may be attributed to artificial selection. A go(
house dog barks at intruders. Barking in all wild canids similai
serves as a warning or threat to intruders (except during grot
vocalization) and is context specific. Is the lack of context specifici
in the dog a hypertrophy of domestication, or does it point to a ve
different nonwolf canid ancestry? Surely it would be of little adva
tage (to man) to selectively breed dogs that bark with virtually i
context-related specificity, although it would be most difficult
exercise context-specific selection. An alternative view is that tl
selection pressure for silence necessary in a wild predator has bei
relaxed. This, coupled with a high degree of dependency and var
ing degrees of neotenization or infantilism with care-soliciting b
havior and vocalizations in adulthood directed toward the owne
would suggest that both domestication and socialization early in li
have a profound influence on the dog's vocal repertoire. The oi
gin(s) of the dog therefore still remains an enigma although oi
might conclude on the basis of this study that if the wolf were tl
sole progenitor of the dog, then dogs would howl more and ba:
much less than they do. Dogs also yelp in many different contex
compared to other canids. TTie prevalence of this sound may ind
cate a clear taxonomic difference between species. Another accep
able interpretation is that the yelp, an infantile sound, persists in
maturity and occurs frequently in distress/attention-seeking coi
texts in dogs because of the dependency promoting neoteny i
infantilism induced through domestication (since the more depei
dent the dog is, the more trainable it is) (Fox, 1972).

Summary

On the basis of speclrographic evidence, it has been possible I
identify 12 basic vocal sound types of canid species.

Vocalizations may be mixed by successive emission of two <
more sound types, by superimposition of these sounds, or by
combination of these two forms. The same basic sound type ma
differ among canid species along the dimensions of sound dur;
lion, separation time between consecutive sounds, principal frt
quencies, cyclicity, and context.

Vocalizations in Wild Canids

89

Developmental data indicate that domestic dogs first begin to
mix sounds by successive sound emissions at about 10 days of age
and later by supeiimposition between 2 and 3 weeks of age. The
frequency of occurrence of the basic sound types in different con-
texts varied between species but not within species. The possible
effects of domestication on canid vocalizations are discussed.

V

Behavior Genetics of
Fi and ¥2 Coyote x
Dog Hybrids

Introduction

This study represents an interim review of the behavior and
morphological features of F , and Fj coyote x beagle hybrids. Earlier
studies by Silver and Silver (1969) and Mengel (1971) are lacking in
detailed analysis of the behavior of parent stock, notably of the
fixed action patterns which to the ethologist can have as much
taxonomic value as morphological features to the comparative
anatomist. Several species-typical action patterns have been iden-
tified for the coyote (Cam's lalmns) and for the beagle (Cams
familiaris) in our laboratory and quantified ontogenetically in sev-
eral animals (Fox, 1969a and b, 1970, 1971a and b; Fox and Clark, 1971;
Bekoff, unpublished observations). The purpose of this section is
to present an integrated overview and to discusss some intriguing
findings which have not as yet been considered.

Materials and
Methods

Developmental data were obtained from 4 male and 4 female
beagles, 6 female and 4 male coyotes, and 16 F. coyote x beagle

9 '

9 ^

THE dog: its domestication and behavior

hybrids, 6 of which were females; of this group, which comprised
six litters, 2 males died prior to weaning. Two male and two female
F, hybrids were obtained as adults and no ontogenetic data were
available for these animals which were bred by artificial insemina-
tion (courtesy of Dr. J. J. Kennelley, Bureau of Wildlife and
Fisheries, Denver, Colorado). All subjects studied developmentally
were hand-raised by stomach tube on Esbilac®, a synthetic bitch's
milk formula (Borden Company). The various morphological and
behavioral characteristics listed in Tables I and II were looked for.
Much of the behavioral data were collected from dyads interacting
between 21-50 days of age and subsequently from groups of ani-
mals housed in groups of three or four. The most significant find-
ings to date are reviewed (see Figure 1).

Results

MORPHOLOGICAL CHARACTERISTICS

All Fi hybrids had flop or pendulous ears, bushy coyotelike tails
with an active supracaudal gland. Dewclaws were absent. The body
type of the F ,'s is best described as mesomorphic, being stockier
than the coyote, but lighter and leggier than the beagle. All F,
hybrids had a smooth coat with guard hairs slightly longer than the
beagle but shorter than the coyote; the winter underhair was much
thicker than in the beagle. (See Table 1 and Figure 1.) It should be
added that two of the Fj canids, both with piebald coloration, had a
medium (coyote) length coat that was broken rather than smooth.
The remaining 12 Fj hybrids had coats intermediate in length and
with underfur density between coyote and beagle, but all coats
were smooth like the beagle (Table I). Their coat color varied, being
dark sable in the winter and paler in the summer.

The Fj generation showed independent segregation of beagle
and coyote characteristics (see Table I). Erect ears, a coyote charac-
teristic, were less frequently seen than pendulous or semipendu-
lous ears. All animals had fairly bushy tails and two had dewclaws
on the hind legs. The piebald coat color (white and brown/black) of
the beagles was seen in a low percentage of Fj's, most of the ani-

Table I.

FREQUENCIES OF MORPHOLOGICAL CHARACTERISTICS OF BEAGLES, COYOTES, F, AND Fj
HYBRIDS

Behavior Genetics of Coyote x Dog Hybrids

93

S3UXU03 /uojoduof

idddn Jo uoijdruj

9d/U'Oi{djoy,f

pm

fZ JZ ^ ^ c

I B- ^ S' S

r*. o w o ^

(iJ 6 S £ S

JO/W }POO
3}auo9u jyjpQ

(s8a] puiii) SrtiopflWQ
puviS ppi patfoy
iwt /fi/sng
s /03 dojj
S /03 P 3 J 3 -IIM 35

&

o i 5

-fi

sp^fqns Jo joquintq

S'

o

•Sc.iwnal v.inations from pale lo dark coat color.
PiRures denote numbers amma^s.

Behavior Genetics of Coyote x Dog Hybrids

95

mals being either pale or dark sable without any large areas of
white on the body, although most F^'s had white around the muz-
zle and lower jaw (i.e., white facial mask of the coyote) and also on
the tail tip. The dark sable type tended to develop a pale coat in the
summer months like the F, hybrids (see Figure 2). The most fre-
quent somatotype was a long-legged coyotelike ectomorph, al-
though a smaller percentage of endomorphs, which were stockier
and sometimes shorter-legged than the beagles, were seen. The tail
gland odor was harder to detect and seemed less active in the
beaglelike Fj's compared to their ectomorph littermates. These
endomorph-type hybrids also had an overwhelming "doggy"
smell, much stronger even than in beagles.

Eruption of the upper temporary canine teeth in the Fj canids
was usually earlier than in the beagles, the age of eruption being
extremely early in the coyote compared to the domestic dog
(Table 1).

One of the most intriguing morphological characteristics was
the neonatal coat color, the significance of which is discussed sub-
sequently. All wild canids are dark at birth (Fox, 1971b) and do not
possess the adult coat color, while most domestic dogs show the
adult color at birth. The majority of the F2 hybrids had a dark
dusty-brown neonatal color, which eventually gave way to a pale
or dark sable. Two animals (Table I) had the adult color at birth,
namely the beaglelike piebald coloring.

BEHAVIORAL CHARACTERISTICS

Both the F, and Fj hybrids were timid and shy of strangers. The
endomorphic beaglelike Fj's were less timid than the more ec-
tomorphic coyotelike Fj's. (See Figures 3 and 4.) Coyotes, even after
being hand-reared, become increasingly shy of strangers with in-
creasing maturity (Fox, 1971b). The wildness of these canids, or
capacity to regress to a feral state, was acddentally evaluated at our
field station. Two Fj hybrids and one beagle escaped from their
enclosure; the latter animal stayed close to the enclosure and was
easily caught, while the hybrids, both ectomorphic or coyote types,
roamed the 2000-acre station for several days before hunger
brought them back and they could be coaxed back into the pen

Behavior Genetics of Coyote x Dog Hybrids

97

with food. Three similarly raised coyotes of the same age were re*-
leased at the field station and were provided with food at the spot
where they were released. They moved away permanently from this
site after 4 days, and were not seen again.

As in wild canids, the first signs of shyness toward novel
stimuli (environmental fear) began to develop around 5 months of
age in a high percentage of the F2 hybrids. The wild temperament
trait is discussed in detail earlier (Fox, 1971b).

Both coyotes and Fj hybrids showed marked aggression to-
ward members of the same sex (Table II). Fj hybrids also showed
increasing proximity intolerance for members of the same sex, al-
though unlike the coyotes (Fox and Clark, 1971), they did not fight
seriously between 4 and 5 weeks of age. Dominance fights oc-
curred later, around 6-9 weeks, and permanent hierarchies were
formed shortly after this age. By 6-8 months, intrasexual aggres-
sion was marked and was the cause of death of one subordinate
male F* hybrid that for necessity had to be housed with male and
female littcrmates.

Clasping a conspecific around the waist during aggre‘;sive in-
teractions (Fox, 1969b), a typical action pattern occa'^ionally ob-
served in the beagle, was fret]uently seen m lxitli_F, and F, hvbnds

Figure 3. Agonistic interactions in coyotes and Fj coyote X
beagles, (a) Intense intraspecific aggression and high proximity in-
tolerance in 29-day-oId coyotes; (b and e) defensive threat gape in
coyote and hybrid; (c and d) more submissive-defensive gape in Fj
hybrid, similar to coyote (ft. See facing page.

Aggressive displays in the coyote include horizontal extension of
the tail, while in the beagle it is arched in a more vertical position
(Table II). In the Fi hybrids, both tail positions were seen, and often
a compromise tail position, between the vertical and horizontal, was
observed during agonistic encounters. F* hybrids showed both
coyote, beagle, and intermediate typo tail positions, the vertical
position occurring in the beaglelikc endomorphs and the horizon-

tal position in the ectomorphic types. The latter hybrids more often
showed the intermediate position than the former, and the tail at
rest was held down as in the coyote.

Threat-gapes (Fox, 1970) and hip-slams (Fox, 1969b) are action
patterns common to the coyote during agonistic interactions and
have not been observed in beagles. Both actions were seen occa-
sionally in the F, hybrids at lower amplitude and frequenc)'; F;
hybrids showed a greater range of variation in occurrence of the
threat-gape (Table II), but all showed the hip-slam during fighting
and play-fighting.

The prey-securing forelimb stab and ability to catch. Kill, dis-
sect, and ingest live prey (4-week-oId rats) will be quantifed in
detail in Chapter 6. Most F. hybrids exhibited the forelimb stab of
the wild canid, and the majority of these animals were efficient at
killing and ingesting prey. Those that failed (interestinglv. the

Figure 4. Action fxittcnis typical of coyote in F* coyote X dog hy-
brids: (A) standing-over, (B) threat-gape. (C) inguinal response
(hind leg is raised).

bcaglct>T3e endomorphs) in executing a normal coyote-killing se-
quence showed inhibition of the prey-killing bite, so that injured
prey were eaten alive. In coyotes, no bite inhibition was observed,
while in beagles, bite inhibition, plus inhibition of eating, w'cre
recorded.

Tnble II.

OCCURRENCE OF OBSERVED BEHAVIORS IN COVOTES, BEAGLES, Fj AND F2 HYBRIDS

Behavior Genetics of Coyote x Dog Hybrids

101

StUfSJ
3SUOdS?J IVUItlSuf

s fv

qpis quiji^joj

U4{fi$-difq

iwi 9<itss^jSv

JVfUOZUOjq

}ivi MissyjSSv

/vjifJM

StndiPiy

MtsspjSSy

uoissjjSSo

pnxpspjiui

fimupjMiiujfi ^ 1/5
sp.flqfis /<»

102

THE dog: its domestication and behavior

The inguinal response, a display of considerable ontogenetic
and social significance in Canidae (Fox, igyid), occurs rarely in
beagles but consistently during social interaction in coyotes. This
display was seen in both Fi and hybrids (Table II).

During friendly sodal approach, the more social canids, such
as dogs and wolves, show a licking intention, often combined with
a submissive "grin" (Fox, 1970). In coyotes, this licking intention is
rarely seen and at low frequency. A slightly higher frequency was
noted in the Fi hybrids and in some individual Fj's; in the latter
there was no apparent correlation with the somatotype.

Estrus in the coyote occurs in early spring (February) so that
the birth of cubs is timed with a relative abundance of prey. In the
beagle, there are two heat cycles per year, usually in the spring and
fall (but in many individuals at alrnost any month of the year).
Estrus in the F, hybrids was restricted to late fall-early winter. The
Fj hybrids were more variable, estrus occurring at any time from late
fall to spring; only one heat was recorded during the summer. One Fj
animal had two heats (spring and late fall), but no fertile mating
occurred.

As early as 3 weeks of age, a marked difference in response to
handling was seen in Fj hybrids. Some pups felt more tense and
had a greater muscle tone characteristic of infant coyotes, while
others remained relaxed, almost flaccid, when picked up, like beagle
pups. There was no correlation vrith these two types and their
somatotype. The more tense pups were more easily alarmed by
sudden noises and would invariably give an open-mouth threat-
gape which is first seen at this age and similarly evoked in coyotes'
(Fox, 1970). An effective, although unsophisticated, test for pain-
induced aggression first used in studying individual differences in
coyote cubs was used on these hybrids. The test consisted of hold-
ing the cub in one hand, abdomen uppermost, and with the other
hand alternately and mildly pinching the cheek skin. In the relaxed
hybrids, there was either distress vocalization or no response. The
tense hybrids reacted like coyote cubs ivith defensive threat-gapes
and at a later age (4 weeks) with growls and vertical retraction of
the lips. Repeated stimulation (10-15 times) would produce a
ragclikc reaction and self-directed aggression where the subject
would seize and bite one or both of its forepaws. This had a vicious
circle effect, self-inflicted pain increasing the aggression. Movies
taken of these two different sets of reactions dramatically illustrate

Behavior Genetics of Coyote x Dog Hybrids

103

this phenomenon, which may be attributed to threshold dif-
ferences in responsiveness to mild cutaneous stimulation. One of
the tense hybrids at 4 weeks of age had to be forcibly restrained
and the canine teeth clipped, after a brief fight with a conspedfic
which triggered self-directed biting; it was unable to walk for sev-
eral days, both forelimbs being severely mutilated. No such ex-
treme reactions (low threshold?) have been observed in coyote
cubs.

The pups identified as tense individuals at 3-4 weeks were
generally more efficient in handling live prey, showed less prox-
imity tolerance and greater intrasexual aggression, and were actu-
ally dominant over their relaxed conspecifics. Future studies will
attempt to quantify such early individual differences in respon-
siveness to cutaneous stimulation.

ADDITIONAL OBSERVATIONS

The majority of F2 hybrids had a characteristic defect in that the tip
of the tongue often protruded out of the mouth while at rest.
Measurements have not been taken, but this may be either a neuro-
logicaJ defect or a disproportion between tongue and muzzle
length. When alert, all but z of the 14 F2 hybrids could completely
retract the tongue. This anomaly was not seen in the Fj hybrids.

The endomorphic F2 hybrids also showed varying degrees of
prognathism, the lower jaw being shortened. Dentition appeared
normal and in no instance was the malocclusion severe enough to
interfere with the alignment of the permanent teeth.

All endomorph beagletype Fj hybrids had a delayed closure of
the parietal fontanelle (4-6 weeks). Like beagles, these latter ani-
mals were more vocal than the ectomorph type. The Fj's had simi-
lar vocalizations, namely a bark-howl which, as in the F, hybrids,
sounded like a successively combined beagle and coyote call.

Discussion

The morphological and behavioral data strongly suggest that the
follovving traits have a dominant mode of inheritance: bushy tail.

104

THE DOG; ITS DOMESTICATION AND BEHAVIOR

pendulous ears, sable wild color, short guard hairs and smooth
coat, seasonal change in coat color, active supracaudal gland, dark
neonatal coat color, coyote while facial mask (around lower jaw
and muzzle), coyote ectomorph somatotype, one annual estrus,
intrasexual aggresslvity, aggressive clasping, hip-slam, threat-
gape, inguinal response.

The dark neonatal coat color, a characteristic of all wild canid
species so far studied (Fox, 1971b), may be related to heat conserva-
tion since neonates are partially poikilothermic. Most domestic
breeds of dog have the same coal color at birth that they will have
as adults, though a greater number show a seasonal change in coat
color.

Eruption of the upper temporary canine teeth, which is very
early in wild canids compared to all domestic species so far studied,
was earlier in both F, and Fj hybrids. As in the beagle, erup-
tion of the lower canines occurred later in the Fj hybrids, and the
upper incisors erupted earlier than the lower, the central pair usu-
ally being the last to emerge. As would be predicted on the basis of
independent segregation of the Fj's into beagle meso/endomorphic
and coyote (ectomorphic) types, tooth eruption was earlier in the
coyote type.

Mengel (1971) also noted a greater variation in Fj hybrids from
doglike to coyotelike compared to the more intermediate F, hybrids.
However, he did not report any detailed developmental data, but
it is significant that all his hybrids were aggressive and fought a
great deal early in life. A similar increase in aggression, compared
to the beagle, was apparent in all Fi and F^ hybrids in the present
study.

Mengel reported small litters from the F, hybrids, and this
corroborates the present findings where the average litter size was
2.7. This may be indicative of a lower fertility in these hybrids.

Gier (1975) and Mengel (1971) have both remarked that shift in
timing of estrus in coyote x dog hybrids would prejudice survival
of any offspring bom in the wild during the winter (except in
temperate southern regions of the United States). These data add
further weight to their argument. Kennelly and Roberts (1969), in a
study of the F, hybrids used in this study, found evidence for
seasonal spermatogenesis in the males.

The inguinal response, a highly ritualized social display in
coyotes which has a clear ontogenetic history (Fox, (1971b), was also

Behavior Genetics of Coyote x Dog Hybrids

105

seen at high frequency in the Fj hybrids but only rarely in beagles.
Its function early in life may be to truncate agonistic interactions
prior to the establishment of social relationships. The high fre-
quency of this display correlates with the great incidence of fight-
ing during the first 3-5 weeks of life in both coyotes and Fj hybrids
compared with the beagle.

Behavioral and developmental studies of these canid hybrids
have made it possible to identify a number of behaviors which, as
emphasized earlier, may be as valuable as morphological traits in
facilitating the identification of patterns of inheritance and of pos-
sible ancestry. Of particular interest and a focus for future research
on subsequent generations and back-crosses is the ontogeny of ago-
nistic behaviors and quantitative analysis of various displays and
action patterns. The canid family holds a wealth of possibilities for
further studies of behavior genetics since many spedes (wolves, dogs,
coyotes, and jackals) (Gray, 1954; Kolenosky, 1971) will hybridize,
and for each of these spedes, characteristic action patterns and
displays have been identified and detailed ethograms and devel-
opmental data are available (Fox, 1971b).

Summary

The frequency of occurrence of various morphological and
havioral characteristics are documented in coyotes, beagles, an
their F, and Fz hybrids. These data reveal certain consistences in
the patterns of inheritance of such beagle traits as pendulous ears,
smooth, short coat; and coyote traits, such as ectomorp y, sa e
coat color, dark neonatal coat color, bushy tail with active supra-
caudal gland, and white facial mask. The above traits were most

frequently seen in the F2 hybrids, while the Fi hybn ^ ^

intermediate between the dog and coyote and extreme T
in appearance. In the F^ hybrids, independent
apparent, most offspring being coyolelike ectomorp ® ^

dulous or semipendulous ears, while others were S
domorphs, but with proportionately shorter legs an f
than actual beagles. Both F. and F= hybrids resembled the coj me m
terms of shyness, intrasexual aggression, an interaction

action patterns associated avith aggression and soci.

VI

Effects of
Domestication on
Prey-Catching and
Killin g in Domestic
and Wild Canids and
F2 Hybrids

Introduction

Following earlier developmental studies of prey-catching and kill-
ing behavior in various species of Canidae (Fox, (1969a), the pres-
ent investigation focuses upon differences in organization an
temporal ordering of these behavior patterns in coyotes, beag^s,
and F, and F, generation coyote x beagle ("coydog ) hybnds The
objective was to determine what effect, if any, dornestication nuiy
have on prey killing in wild and domestic canids an t cir )

This investigation is dii-idcd into two parts, the first dea mg
with more qualitative aspects of behavioral organization and em-
poral ordering or sequencing and the ^econ "i u

malysis of two action patterns, the vertical leap and the forehmb
stab. (For detailed descriptions of these action pattims, see ■ .

1969 and 1971b.)

107

io8 THE dog: its domestication and behavior

Materials and
Methods

PREY-CATCHING AND KILLING

In the qualitative part of this investigation, subects were 8-9 weeks
old when tested and had no prior experience with live or dead
prey. Differences in behavior might therefore be attributable to
genetic rather than experiential influences. The temporal ordering
of action patterns and the frequency of certain action patterns were
recorded for qualitative and quantitative analysis, and tests were
repeated in order to determine what effect experience might have
on subsequent prey-catching and killing behavior. Additional
studies were made on six coydogs in order to evaluate the possible
Influence of experience in the test/retest design of the experiment.

Four beagles (littermates), 9 Fj (from 3 litters), and 8 Fj (from 3
litters) generation beagle x coyote hybrids and 13 coyotes (from 3
litters) aged 8-9 weeks were used in this study. They were hand-
raised in incubators from 3-6 days of age, being fed regularly on
Esbilac via stomach tube, as described by Fox (1966). This method
was chosen to control for individual differences in maternal be-
havior. All subjects were raised indoors in pairs or trios and had no
exposure to live or dead prey prior to testing. They were never fed
fresh meat and were raised on commercial dog food (Purina Chow
and Gaines Burgers) being weaned at 3 weeks of age onto this
food blended with Esbilac and housed in littermate groups in 4
X 10 ft cages. Each subject was carried from the animal room into
the soundproof testing room opposite and was then observed
alone in a 5 X 5 ft arena, where the prey, a 4 to 5-week-old white
rat, was placed. This type of prey was chosen for convenience,
being readily available and having been used consistently in earlier
studies with canids. An arbitrary cutoff time of 20 min was as-
signed for each test, during which all behavior was recorded
against a time-base using world-symbols and abbreviations for the
various action patterns (see Tables I-III). It was then possible to
quantify the frequency of certain action patterns and to trace tem-
poral sequences of prey-catching, killing, dissection, ingestion.

Effects of Domestication on Prey-Catching and Killing

109

and play with prey. If the subject did not kill and/or eat the prey
during this 20 min period, the prey was killed immediately after
the test and was presented to ^e subject with the abdomen cut
open to expose the liver and viscera, and observations continued
for a maximum period of an additional 30 min. If the subject gave
no response after 5 min, the liver was removed and placed in its
mouth in order to sensitize the subject to the carcass. This is occa-
sionally effective in stimulating a naive canid to eat prey for the
first time (Fox, 1969). The action patterns associated with dissection
and ingestion of the prey were recorded until the prey was eaten or
partially eaten and ignored. This time period varied up to approx-
imately 30 min.

All subjects were retested 24 hr later in order to determine
what effects the first experience with prey had on subsequent or-
ganization and sequencing of action patterns.

VERTICAL LEAP AND FORELIMB STAB

From a qualita«ve study of prey-catching and kiUing in coyotes,
beagles, and F2 generation coyote x beagle hybrids, it was observed
that two very distinct action patterns were a consistent part of all
three of the above species' behavior repertoire. These action pat-
terns are the forelimb stab and the foreUmb stab with a vertical leap
(see Figure 1). The latter is analogous to pinning down descnbed by
Eisenberg and Leyhausen (1972) in other predators.

It was felt that quanHfication of the frequency of the occur-
rence of these action patterns in controlled observational situations
might render an appropriate index of the degree to which these
three species were aroused and moHvated to e lat these action
pattern! The second section of this chapter, therefore deals specif-
ically with quantified analyses of these particular behaviors.

After completing the prey-kilUng tests at 9 u-eeks of age, t
same litter of four beagles, hvo of the litters of six coydog hy-
brids, and one of the litters of three coyotes, were kept for this
study. Four beagles (littermales), six F, generation beag e x co> o c
hybrids (littermates), and three coyotes (also littcmatcsk . 5 9 -

weeks of age were thus used in this study. Each subject uas ob-
served in the 5 X 5 ft .arena, the floor of which ivas covered bv a

110

THE dog: its domestication and behavior

Figure i. Vertical leap (a) followed by forelimb stab (b) directed at
visually concealed prey by generation coyote X dog.

Effects of Domestication on Prey-Catching and Killing

111

white linen sheet. The edges of the sheet were secured under the
plywood walls of the arena. The prey, two 4 to 5-week-old white
rats, were placed in the arena under the sheet. The two action pat-
terns (vertical leap and forelimb stab) were most often elicited
when the canids observed the movement of the rats under the
sheet. Each test was arbitrarily divided into two 10-min intervals
(20 min for each test), during which the frequencies of these two
action patterns were recorded. The total interaction time between
canid and prey was also recorded. In addition, qualitative notes
describing major events such as death of the prey and overt signs
of degree of arousal of the subjects were also recorded. Each
20-min test was repeated within a week after the initial test.

Results

PREY-CATCHING AND KILLING

Beagles

In the four beagles tested, all subjects reacted toward the rat;
but interactions were brief, and the rat would frequently be ig-
nored as the dog suddenly began to investigate the arena or at-
tempted to get out of the arena. These observafaons clearly imply
that the beagles were less motivated than the coyotes and coydogs.
No prey was kiUed in either the first or second tests.

The following behaviors were recorded: onentation, approach,
follow, investigation (sniffing and pawing), approac wi '

bite intention, Lse-stab, head-shake intenhon leap-leap foHowed
by bite intention, nose-stab or head-shake intenhon P^mg
■aese actions may be followed by an ,he

sideways. Only one beagle contacted the live rat with
prey being licked and gently nibbled with the incisors. All subjects
showed plav-soUciting behaviors, including the bow, exaggerated

and also barking and yelping

■' '"LIL ....... «»

patterns were observed, bite, carry, one

112

THE dog: its domestication and behavior

forelimbs, hold with one or both paws and bite. No dog killed the
rat in the first or second tests.

These observations show that although the dogs were at-
tracted toward the rat, the duration of interaction (and therefore
possibly the motivation) was much shorter than in coyote or
coydog and a number of prey-catching and killing action patterns
were absent. The normal canid temporal sequence of orientation,
approach, investigation, bite, carry, bite (and head-shake) to kill,
followed by dissection and ingestion was truncated at the bite in-
terphase. Investigation was instead followed by self-play (chasing
own tail), play-soliciting (which is not unusual in wild canids but
which is normally satiated after 10-15 min), or by a bite intention.
The dogs would lunge at the rat and bite in vacuo (air-snap) or
deliver nose-stabs which were interpreted as bite intentions. Such
actions were often preceded by forward leaps and followed by one
or more backward leaps. This leap-leap-nose-stab sequence (see
Table I) was only observed in the beagles and also occurred during
intraspecific play. The same sequence of actions has been recorded
in Chihuahuas, malemutes, malemute x wolf hybrids, and several
domestic dogs of mbced breeding, but not in coyotes or wolves
during intraspecific play. The possibility that this action pattern
sequence is a specific marker lor Cams familiaris, as a species charac-
teristic, should be considered.*

When retested with live prey after 24 hr deprivation of food,
only one of the four subjects showed no additional inclusions of
the actions associated with the prey-catching and killing sequence.
In one subject, the rat was bitten 13 times, and on each occasion the
bite was inhibited, the rat being uninjured and active at the end of
the test. The rat was grabbed and carried five times by this subject,
and on the third carrying episode (15 min after onset of test) a weak
head-shake occurred which was more violent on the fourth carry-
ing episode but did not occur subsequently. This subject also made
four low amplitude forelimb stabs at the rat during this second lest.
In the first test, this subject directed neither biles nor forelimb stabs
at the rat. Another subject added only two inhibited bites to its
earlier repertoire.

•More important though is the fact that this is an Inlraspcdfic action that was
directed at the prey which, as cmphastied earlier (Fox, 1969), may be responded to
as a play*objcct or partner in some canids until it is learned that the prx’y is food.

Effects of Domestication on Prey-Catching and Killing

113

Table I.

EXAMPLE OF INDIVIDUAL NOTATIONS

Beagle Male II (First Test)

No immediate orientation.

2 min

Then orients, rushes, paws, TW (tail-wag), sniffs, then ignores,
then approaches, TW, bite intention, head-shake intention.
Self-play leap-leap-leap-nose-stab.

5 min

Investigates then ignores rat.

6 min

Leap-leap-nose-stab, paws, chases, bite intention, investigates,
TW.

7 min

Licks rat's face.

Paws, bite intention, play-solicits, TW.

8 min

Lunge, nose-stab, A/W (approach/withdrawal), play-solidts.
Nose-stab, A/W, bite intention.

20 min

Play-solicit, bite intention, yelp-growl, TW.

(Brief poorly sustained interactions — no clear bite, cany or
forelimb stab.)

22 min

Self-play.

12 min

Rushes, gently nibbles, paws, A/W, nose-stab, bite intention.

TW, play-solicit, paws, yelps, leap-leap-A/W nose-stab, bite
intention, yelp.

14 min

A/W, bite intention, TW.

Leap-leap-leap-nose-stab, A/W, play-solicits.

Leap-leap-leap away, bite intention, leap away.

15 min

Growl-howl-bark at rat. Play-soliciting dance— sustained playful
interaction for last 5 min.

(Rat is alive and well.)

Beagle Female IJ (Retest)

Immediate rush, paws, and iw/es, but bite inhibited. Then ignores
rat.

2 min TW, bite — carry, then dropped in one comer.

Inhibited bite, then bite intention.

Yowl-barks at rat + nose-stab.

114

THE dog: its domestication AND BEHAVIOR

Table I. {continued)

5 viin Approaches and mouths and leaps away.

Leap-leap and mouths.

7 min Mouths, leaps back, then bile intention.

Plays with tail of rat.

Scif-play.

Play-soUdts + nose stab.

n min Leap bile (inhibited).

Paws and mouths.

15 min Ignores rat, little interest.

16 /nin Carries, drops, one forelimb slab.

Carries, weak head-shake.

Carries, strong head-shake.

Drops, mouths, carries, mouths, drops, /orelimb stab.

28 min Stalks and rushes, forelimb stab, growl.

Bite intention.

(Rat is alive and well.)

A third subject showed a marked increment of actions, giving
14 inhibited bites and 8 forelimb stabs and grabbing and carrying
the rat 10 times. On the fourth carrying episode, piloercction oc-
curred, and on the sixth episode (10 min after start of test) the first
of three head-shakes, which subsequently increased in intensity,
were noted.

An interesting and bizarre coupling of action patterns was
recorded after the second grab-carry episode, which occurred 8
times. This was the inclusion of a high amplitude leap, which on
the last three occurrences, was coupled with a head-shake but
which occurred after the dog had leaped at, grabbed, and carried
the rat. Instead of the usual temporal sequence of leap, grab and
bite, head-shake and can)* (or cany and head-shake), this subject
coupled the carr)* component with jin additional leap while it had
the rat in its mouth. And on three later occasions when this mis-
placed leap was seen, a violent head-shake was recorded simul-
taneously.

Effects of Domestication on Prey-Catching and Killing

115

It is logical to propose that this prey-catching leap (while the
prey was in the dog's mouth) was a misplaced component of a
disordered temporal sequence and that it subsequently entrained
the head-shake component which normally occurs after a leap
when the prey is secured in the dog's mouth.

When given dead prey with the viscera exposed after the first
test, additional evidence of the dogs' inability to efficiently dissect
the prey was obtained. Only one of the four subjects ignored the
carcass completely. A second subject ignored the cut-open rat until
the rat's liver had been placed in its mouth. This was ingested and
the dog then immediately went over to the carcass and licked out
the viscera. Excessive licking continued for over 4 min. Ths was
followed by the dog repeatedly lifting up the carcass with the in-
cisors; the forelimbs were not used to secure the carcass beneath
the feet to facilitate dissection. This lack of coordination of
forelimbs and teeth continued for approximately 6 min, when the
dog lay down with the rat alternately between and under the
forepaws as it indiscriminately chewed and mouthed the abdomen
and appendages of the carcass. After 15 min, one forepaw was
severed. Each time the pup pulled on the carcass, the latter would
slip from under a forepaw. After 25 min, the head was removed
and ingested.

A third subject immediately grabbed and shook the carcass,
gave several crush-bites, and attempted to bury it in one comer of
the arena. This subject ate the rat's liver when given it ut su
sequently ignored the carcass. The fourth subject also ate the Ever

but ignored the carcass. , , . _

After the second prey-killing test, all four s"bl<=cts were mor
reactive toward the carcass. The first subject bcked ou
and inappropriately used the camassial teeth to ^ ,1,^

viscera. This took over 15 min. The normal reaction is to tear the
viscera with the incisor teeth once the prey has

the forefeet for counter-traction. This subjert never

under the forefeet and subsequently ignored rat after ingesting

the liver; it appeared incapable of dissecting t

The second subject violently shook the ^

then appropriately (Lshed and severed the head
sials and ingested it. The rest of the carcass was masticated ivith

ii6

THE DOG: ITS DOMESTICATION AND BEHAVIOR

the camassials, the forelimbs now being better integrated with
teeth actions, in that the subject would frequently hold the carcass
under one paw while chewing the rat with the camassial teeth.

Instead of using the incisor teeth and counter-traction with the
forepaws to tear the skin, limbs, and viscera, the third subject
continued to masticate the carcass with the camassials. Con-
sequently, the carcass was not dissected or eaten.

The fourth subject first used the camassials like the preceding
subject, but after 6 min switched and used the incisor teeth in
combination with forelimb counter-traction to tear and dissect the
rat. The viscera were ingested followed by a long period of chew-
ing the rat's appendages with the camassial teeth. This latter inef-
fectual activity after 8 min was directed to the head of the carcass,
which was cmshed, severed, and ingested. The remainder of the
carcass was subsequently ingested. These observations are pre-
sented in detail in order to demonstrate the role of experience in
improving prey dissection and ingestion.

Coyotes

The detailed ontogeny of prey-killing behavior in the coyote
has been described earlier (Fox, 1969), and only the major points
relevant to the present investigation will be presented (see Table
11 ).

In one Utter of four 8-week-old coyotes, all immediately
grabbed and bit the rat, the bite being uninhibited. The coyotes
immediately killed the rat and had dissected and eaten it within 1
and 3 min, respectively. The other two killed their prey after 2.5
and 5 min, respectively, each taking a further 2.5 and 7.5 min to
dissect and eat the carcass. The slowest of these four subjects en-
gaged in excessive bouts of head-shaking which was released
every time the prey moved in the coyote's mouth. Long bouts of
violent head-shaking continued during dissection of prey; when
later retested, this same coyote engaged again in bouts of excessive
head-shaking, but the bouts were less frequent than on the first
test. All subjects when retested showed improved orientation of
the grab-bite, striking the prey in the anterior thoracic region rather
than aiming more posteriorly, which would give the rat sufficient
freedom of movement to bite the subject's face. When retested all
four subjects had killed and ingested the prey within 60 sec (aver-

Effects of Domestication on Prey-Catching and Killing

117

Table II.

Coyote Female II, Group II (First Test)

Immediate grab-bite-carry, then crush bite. Was bitten on

1.5 min face— released rat, then grabbed around neck and thorax +

violent head shakes.

Multiple crush bites.

2.5 min Crushes TW, AAV, TW, crush bites, AAV, TW. Crush bites,

carries. AAV, TW, carries— intermittent play and crushing.

4.5 min Pulverizing, holds with forepaw.

Crushes head with camassials, holding body with one forepaw.

7.5 min Eats off hind leg, then tail.

11 min Eats off other hind leg. Uses camassials and forelegs well.
Rips skin with incisors using forelimb counter-traction.

33 min Eats intestines.

24 min Uses camassials and removes and eats head.

age 35 sec). None of these four coyotes played with the live or dead
prey. All showed perfect coordinaHon of foreUmbs and teeth m
pulling the carcass to pieces and used their teeth appropnalely
the camassials to shear and crush and the inasors to puU and ear,
in contrast to the beagles described earlier (that lacked fo^h^
coordination and used their teeth inappropnately, such as using
the camassials to tear the skin and viscera).

In a second Utter of six coyotes, the average killing tune w as
3.3 min, all subjects immediately orienting and
with no bite inhibition. Usually multiple cms . ‘ .

terspersed head-shakes killed the rat. Two

brief play-soliciting and tail-wagging, an a s ,

was inhibited. All subjects threw the rat across he arena wi h .
powerful upwards and sideways head flick, the f'

beine creater in the two subjects that engaged >n bnef bouts ol

preceded by a vertical leap. Onl) one s J
dead prey, even on retest.

ii6

the dog: its domestication and behavior

the camassials, the forellmbs now being better integrated with
teeth actions, in that the subject would frequently hold the carcass
under one paw while chewing the rat with the camassial teeth.

Instead of using the incisor teeth and counter-traction with the
forepaws to tear the skin, limbs, and viscera, the third subject
continued to masticate the carcass with the camassials. Con-
sequently, the carcass was not dissected or eaten.

The fourth subject first used the camassials like the preceding
subject, but after 6 min switched and used the incisor teeth in
combination with forelimb counter-traction to tear and dissect the
rat. The viscera were ingested followed by a long period of chew-
ing the rat's appendages with the camassial teeth. This latter inef-
fectual activity after 8 min was directed to the head of the carcass,
which was cmshed, severed, and ingested. The remainder of the
carcass was subsequently ingested. These observations are pre-
sented in detail in order to demonstrate the role of experience in
improving prey dissection and ingestion.

Coyotes

The detailed ontogeny of prey-killing behavior in the coyote
has been described earlier (Fox, 1969), and only the major points
relevant to the present investigation will be presented (see Table

In one litter of four 8-week-old coyotes, all immediately
grabbed and bit the rat, the bite being uninhibited. The coyotes
immediately killed the rat and had dissected and eaten it within 1
and 3 min, respectively. The other two killed their prey after 2.5
and 5 min, respectively, each taking a further 2.5 and 7.5 min to
dissect and eat the carcass. The slowest of these four subjects en-
gaged in excessive bouts of head-shaking which was released
every time the prey moved in the coyote's mouth. Long bouts of
violent head-shaking continued during dissection of prey; when
later retested, this same coyote engaged again in bouts of excessive
head-shaking, but the bouts were less frequent than on the first
test. All subjects when retested showed improved orientation of
the grab-bite, striking the prey in the anterior thoracic region rather
than aiming more posteriorly, which would give the rat sufficient
freedom of movement to bite the subject's face. When retested all
four subjects had killed and ingested the prey within 60 sec (aver-

Effects of Domestication on Prey-Catching and Killing

119

Table III.

Coydog, "Pale" Male (Retest)

Immediate chase, grab-bite, cany, then investigates. Chases,
carries, holds in paws, TW, bite intention, carries, mouths.

5 min TW, nibbles (no blood). Play-soliciting posture, paws, nibbles,

carries, weak head-shake.

6 min Bite clearly inhibited; play-soliciting and TW combined with

grab-bite, head flick and pawing and carrying. Bile becoming
less inhibited.

8 min Blood, mouthing, holds with forepaws, paws, frequent nibbles,

paws, rat bites pup on Up — distress vocalization. Then more
violent play, head flick, leap, grab and carry.

9 min Holds with forepaws and begins to eat leg. Rat still alive. Now

AAV, TW, play leaps, briefly, then chews and pulls forelimb,

' holding rat down with forelimbs. Rat dying.

13.5 min Now eating rat proper— eats head first.

beagles, interaction with the prey was conHnuous (i.e., as m the
coyote, motivation or arousal was more intense t an in e
beagles). One subject in the first, and this same su jec an
Others in the second test ingested dead prey (presen e wi
abdomen cut open), all three animals showing per ect coor '
of head pulling and forelimb counter-traction to tear e pr ^
incisor and camassial teeth were used appropna e y o ^ ' '

and shear the carcass, and all subjects followe w a 1 ^

be the normal canid pattern of crushing, sevenng, and ingest g

atole observations are summarized - Tabje IV which
serves to illustrate the general trends of the three g p
studied in this investigation. md

In order to controWor the interaction Zs

retest of prey-killing, where ■''f'" group of three

presented with dead ,esied as follows: three

6- week-old F, coyote x dog hybnds were lesic

THE dog: rrs domestication and behavior

118

When retested, all subjects spent less time chewing and sever-
ing the appendages of the prey and assumed the normal sequence
of crushing, severing, and ingesting the head, then crushing and
tearing the rest of the carcass, pieces of which were ingested at
random, usually the pelvic region and hind legs being ingested
last; the caecum was rejected by two of the subjects.*

F2 Coydogs

All coyote x dog hybrids showed varying periods of play with
prey, and all had clearly inhibited bites. Leaping and catching pat-
terns were well coordinated and integrated, but the killing bite was
absent. Most subjects toward the end of the first test, as in the
beagles, developed more intense head-shaking, but in contrast to
the beagles, the grab-bite-carry sequence was present at the begin-
ning of the test. As with the coyotes, these hybrids never vocalized,
while all beagles barked, yelped and growled at the live prey.

The doglike leap and nose-stab was also seen in all six Fa
hybrids. They would throw the live prey across the arena with a
head flick and then leap upon it with a coyotelike high amplitude
leap and forelimb stab. Only one of these subjects killed the rat
after approximately 10 min, and, in fact, no killing-crushing bite
was applied; the rat was essentially eaten alive, and on retest, this
same occurrence was recorded. This clearly underlined the omis-
sion or inhibition in the coydog hybrids of the killing bite, although
all other prey-catching patterns were evident (see Table III).

As the tendency to shake the live prey increased toward the
end of each test and was much more frequent in the second test, so
the inhibited bite and multiple nibble-bites became more intense.
In the second test, only one other coydog killed the prey after 8
min, and the multiple killing bites (combined with violent head-
shakes) appeared to be disinhibited nibble-bites. This subject sub-
sequently ignored the dead prey. The frequency and duration of
play bouts were less in all subjects in the second test, yet unlike the

Ejfects of Domestication on Preif-Catching and Killing

121

Table V.

CHANGES WITH SUCCESSIVE DAILY TRIALS
IN THREE F2 COYDOG HYBRIDS

Response latencies
of subjects (sec)

Trials

Actions patterns

5.“

S,"

s.

I

grab-carry

245

45

325

head-shake

—

—

325

crush-bite

—

—

—

n

grab-carry

30

10

5

head-shake

—

—

5

crush-bile

—

—

—

HI

grab-carry

10

45

15

head-shake

—

360

15

crush-bite

—

380

®Ate dead, dissected prey at end of test.
‘’Partially killed and ate live prey.

was inhibited in all trials and no head-shakes were recorded, al-
though the actions of grab-bite and cany, stalking, leaping, an
play-soliciting were recorded. Occasional head-shake
while approaching the rat were recorded in this subject in na s
and in. The subject handled dead dissected prey in an uncoordi-
nated manner resen ibUng that of the beagle, and the general
phenotypic features of this Fz hybrid \vere closer to beag ^
coyote. S„ which phenotypically resembled rnore 'he
its prey on the thiJd trial after 8 min; the head-shake first appeamd

in this trial after a long latency, and as observed in e o ? • ■

group, there was a^dual disinhibition of

nibble-bites increasing in intensity until the Pr ^

around the thorax. This subject, as also obser^'cd in the f g ^p

of hybrids, began eating the prey while it was ® ‘ ‘

and S 3 showed good coordination of ‘ stronc head-

gesting dead prey. Throughout all trials Ss shou

120

THE dog: its domestication and behavior

Table IV.

normal temporal sequence of prey-catching and killing
in coyote

(With point of truncation ofsequencein beagles o/Fg coyote x dog hybrids)

Orientation ►approach ►investigate ► play

grab-bite (with or without leap
and forelimb stab)

carry

head-shake

2 *>..

crush (kill) bites
- and head shakes

dissect

ingWt"

“1, Level of truncation of temporal sequence in all four beagles, first test.

Level of truncation of temporal sequence in all four beagles, second test.

'3, Level of truncation of temporal sequence in all F* coydogs, first test, the kill-bite
being inhibited in four out of six subjects.

'‘Complete follow-through of temporal sequence occurred in only one coydog, and in
this subject there was no clear kill-bite.

consecutive days of testing, each test lasting lo min, and dead
dissected prey being given not after the first trial as in earlier
studies, but after the second trial day. In this way, the contribution
of experience with live prey in Trial I influencing performances on
Trial U could be separated from the possible added effects of ex-
perience with dead prey between Trials I and II.

The data are summarized in Table V. An overall decrease in
response latency over successive trials is evident. In Si, the bite

Effects of Domestication on Prey-Catching and Killing

123

one individual in the second litter killed the rat after 10 min; multi-
ple nibble-bites to the body became disinhibited and suddenly the
thorax of the prey was crushed. The rat was then eaten. Although
the prior group test in the arena with prey did not reveal any social
facilitation of prey-killing, social facilitation (i.e., competition) may
have triggered prey-killing in the home pen, as emphasized by
Leyhausen (1973) in his studies of prey-kiUing in cats. The signifi-
cance of a particular place to eat and hunt is worth further investi-
gation. Burrows (1968), for example, records that red foxes may not
kill prey within their denning area, and a rabbit or ground-nesting
duck may even share the same denning site. The same animals
would undoubtedly be killed if found in the fox's range when it
was out hunting.

RESULTS OF FORELIMB STAB AND LEAP-STAB
OBSERVATIONS

In the four beagles tested, all subjects interacted with the prey mov-
ing under the sheet. However, the interactions were of very brief
duration, and all four subjects never exhibited either vertical leaps
or forelimb stabs. The quantitative results of all tests are reported in
Table VI. In most trials the beagles spent a great deal of time in
both play-soliciting and self-play behavior. Frequently a subject
would completely ignore the prey for most of the test and then
finally spend a minute or two growling, barking, and nibbling at
the prey. All subjects would claw with their forepaws at the prey
when the prey was lying immobile against the side of the observa-
tion arena. When the prey attempted to escape from the beagle, the
subject's most intense response would be a nose-stab, whereas the
other two species would respond with a forelimb stab or a vertical
leap and stab. There was no significant difference found to exist
between Trials i and 2 for the beagles' scores for forelimb stabs and
forelimb stabs with vertical leap.

In contrast to the beagles, the three coyotes tested showed
high frequencies of both action patterns and on the whole the
coyotes spent significantly more time interacting with the prey. The
coyotes followed every mov'ement of the prey; they would stalk

122

THE dog: its domestication and behavior

shake action, but the bite was inhibited and all prey were unin-
jured at the end ol each trial. Clearly, the experience with dead
prey between Trials II and III had no appreciable effect on the
prey-killing tendencies of Si and S3. Such experience may have
contributed to the development of crush-bites which almost killed
the prey of S3, but another interpretation can be offered, namely,
the increasing arousal, short response latencies and increasing du-
ration of interaction over successive trials and the close temporal
linking between head-shaking and crush-bites in this subject. Fur-
ther studies are needed in order to determine the relative impor-
tance of maturation, increments of experience with live prey, and
the effects of ingesting dead prey on the development of preda-
torial action sequences.

F3 Coydogs

None of the eight F3 generation coyote x dog hybrids from
three litters killed prey on the first test, and when retested, only
one killed and ingested its prey. With this one exception, all sub-
jects showed only brief interest in prey during the first test, reac-
tions being orientation and following but no contact. When re-
tested, additional reactions appeared, grab and carry in two sub-
jects and head-shake and bite intentions in another. Reactions in
the third trial were virtually the same. The one F3 hybrid that killed
after a latency of 6.5 min on the second test was the most respon-
sive in the first test. It immediately grabbed and then carried the
live prey and after 7.5 min gave a mild head-shake. After this it
continued to play intermittently with the prey until the end of the
test period.

Clearly, the F3 hybrids have a significant repression of prey-
catching and killing behavior. This could be because they were
more disturbed than the other canids when in a relatively unfamil-
iar place (i.e., the testing arena). In order to assess this possibility,
they were tested again in their home cages. (The day prior to this,
each litter of three pups was placed together in the arena with live
prey. It was virtually ignored, subjects engaging in social play in-
stead.) In each litter a tug of ivar began over the dead prey placed
in their cage. It was not eaten, however, and was treated like a play
object. With live prey, one litter did not kill in its home pen, while

Effects of Domestication on Prey-Catching and Killing

125

and chase the rats, trap them, and make uninhibited bites, head-
shakes, and several forelimb stabs and leaps. One coyote actually
killed the prey. In the two tests, the prey was killed with a crush-
bite, and then the coyote attempted to pull the dead rats through
the sheet. In three other subsequent retests, the prey were appar-
ently the victims of the forelimb stab with vertical leap. In these
cases, the subject ignored the rats after they were killed.

Among the F, generation beagle x coyote hybrids, there was a
high frequency of both forelimb stabs and vertical leaps, and the
total interaction time between subjects and prey was higher than in
both the beagles and coyotes. (See Table VI.) All of the hybrids
required approximately 2-5 min to adapt to the observation arena.
In contrast, the beagles sometimes spent the first 10 min adapting.
This adaptation period consisted of the following types of be-
havior: initial investigation of the arena (ignoring the prey), several
attempts to escape from the arena, self-play (tad-chasing and
jumping in the air), and finally encounter with the prey, usually by
accident. Once the Fj hybrids did encounter the prey they became
highly aroused. Most subjects exhibited a high frequency of
forelimb stabs. These were usually elicited by small movements of
the prey. In contrast to the coyotes, the beagle x coyote hybrids
never used a crush-bite on the prey (emphasizing this same point
observed in the previous test with live, visible prey). All the hy-
brids' oral contact with the prey consisted of either inhibited biting
or multiple inhibited itibble-bites. As in the case of the coyotes, the
coydogs exhibited a high degree of arousal and motivation as evi-
denced by the high frequencies of forelimb stabs and leap-stabs,
long bouts of intense stalking behavior, and long periods of in-
teraction time with the prey. The hybrids also displayed a high
incidence of what may be referred to as place memory. For exam-
ple, the subject would be interacting with one of the two prey, then
would move to the other rat some distance away and then, after
some time, suddenly turn and make a high ampUtude vertical leap
directly at the spot where the first prey had been located. Many of
the leap-stabs were observed in this context or when one of the
hybrids was chasing a fast escaping prey. In the latter case, the
subject would make one or two fotelimb stabs at the prey; the prey
would then start to flee, at which point the subject would make a
vertical leap and stab.

r.iWf VI.

124

THE dog: its domestication and behavior

Effects of Domestication on Prey-Catching and Killing

127

The motivational state may therefore change with experience
and may vary between individuals and between species and hy-
brids; differences in conflicting motivations, such as the tendency to
play or to be fearful of the test situation as emphasized by Mason
(1967), must also be considered.

It may be postulated, therefore, that optimal arousal and
motivational conditions are essential for the release of certain ac-
tions before experience per se can have any effect upon them.

The inferior prey-killing ability in beagles and in some of the
coyote X dog hybrids may be due to a delaying effect (of domestica-
tion) in the maturation of this behavior. Thus, it may be argued
that the age of 8 weeks may be too early to test prey-killing in
beagles and certain hybrids. Such animals when retested at 10 and
12 weeks of age did not, however, show any further improvement
in prey-killing ability, and it is unlikely, therefore, that there is any
delay in maturation per se, but rather, as implied from this study,
there is a change in threshold and truncation of the temporal or-
ganization and sequencing of action patterns attributable to domes-
tication.

Analysis of the two action patterns, forelimb stab and leap-
stab, reveals that these actions are a predominant part of the be-
havioral repertoire of the coyote and coydog hybrid, but under the
test conditions (concealed prey) they were not reported in the
beagles. This apparent absence in the beagles may be correlated
with their low overall duration of interaction scores, which indi-
cates that they were less aroused than the coyotes and hybrids. The
fact that in one beagle leap-stabs were occasionally recorded when
retested with live visible prey supports the arousal hypothesis al-
luded to earlier in this discussion. A second possibility is that the
forelimb stab and leap-stab occur at a lower frequency in the
beagles because they have a higher threshold; a greater degree of
arousal is needed for its release than in the coyotes and hybrids.

Bite inhibition was clearly manifest in the beagles and their
coyote hybrids, and inhibition could not be attributed to moti-
vational factors alone since the hybrids were highly motivated
even though they showed bite inhibition. This notion is supported
by comparative developmental studies of early contactual and ag-
gressive behaviors in coyotes, wolves, dogs, and wolf X dog hy-
brids (M. W. Fox, personal observations) and prey-killing abilities

126

THE dog: its domestication and behavior

Discussion

The two related studies bring out a number of intriguing points,
the genetics of which can only be speculated upon at this stage in
view of the small sample of coyotes, beagles, and hybrids used in
this project. More precise genetic inferences may be drawn in fu-
ture studies involving not only a larger n but also Fj and back-cross
coyote X dog hybrids. In spite of these limitations, qualitative and
quantitative differences in prey-catching and prey-killing were re-
markably consistent within groups arid add to our understanding
of the effects of domestication on canid behavior.

The most striking and consistent findings were the absence of
bile inhibition in the coyotes in the two tests where the prey was
either visible or visually concealed and complete inhibition of the
prey-killing bite in the beagles. In contrast to these extremes, two
of the Fj hybrids showed a gradual release from inhibition, and
although prey were eaten, they were not as effectively killed as by
the coyotes.

The difference in the degree of bite inhibition may account for
the truncation of temporal sequences of prey-catching and killing,
but it should be remembered since temporal organization is varied
during play, that coyotes will play vrith live prey for variable
periods and at such times show clear bite inhibition prior to killing
the prey (Fox, 1969). The truncation of the sequence may be at-
tributable to differences in arousal in the beagles and hybrids com-
pared to the coyotes, a high degree of arousal being necessary for
actions such as crush-(or kill) bite, head-shake, forelimb stab and
leap-stab to be elidted.

Another related hypothesis is that the action patterns as-
sodated with prey-catching and killing are ordered temporally on
the basis of threshold differences, those actions toward the end of
the sequence having a higher threshold than earlier actions. A
minimally aroused subject would then only show low threshold
responses such as orientation, approach, and follow, as was the
predominant feature of the beagles in the first test. Arousal (or
motivation) may therefore increase with experience with prey, so
that higher threshold action patterns are incorporated appro-
priately into the temporal sequence. This hypothesis is supported
by the retest data from beagles with visible prey.

Effects of Domestication on Prey-Catching and Killing

129

Table VII.

Complete hunting sequence of
wild canid

Tracking, trailing

Herding, driving

Stalking, pointing

Attacking, killing

Retrieving
(for cubs/mate)

Partial sequence of some
domesticated dogs^

Bloodhound, gazehound

Sheep dog

Setter, pointer

Boarhound

Retriever

■Attack inhibited in most breeds.

average kill latency of 15.5 min. When retested, this latency drop-
ped to 10.5 min. Interestingly, one pup, prior to killing, held the
prey with its forepaws and pulled at its legs and tail, the consum-
matory (eating) phase occurring before the killing bite.

This clearly demonstrates that inheritance does influence bite
inhibition — a point of considerable social significance in view of the
number of people being bitten by dogs today (estimated 400,000
per year in the United States). While improper human handUng
may often be to blame and/or inadequate socialization (Fox, 1972), a
relaxarion of selection for stable temperament and bite inhibition
may also be involved especiaUy in those breeds that become popu-
lar and are mass-produced, with Uttle quality control, to meet pubUc
demand. The release from inhibition in well-socialized dogs for
attack training can be effected by increasing arousal through play-
fighting with a padded sleeve. A gradual disinhibition occurs
without disruption of the social bond. This unlearning of a socially
inhibited response (biting) is analogous to the gradual appearance
of the killing bite in coydogs where motivational factors or social
inhibition per se may be involved. The reaction of beagles toward
live prey was essentially one of social play and logically, social
inhibirion of biting occurred. This conclusion was also made m
earlier studies of prey-kilUng behavior m wolf cubs. Those that did

128

THE dog: its domestication and behavior

(Fox, 1969). Around 24-30 days of age, contact with a conspecific
frequently releases an uninhibited bite in coyotes (and also in
young red foxes of this age), and it is only after dominant-
subordinate relationships have been established that the bite be-
comes inhibited. Prior to this time, play behavior is rarely seen. In
contrast, wolves and dogs show marked intraspecific bite inhibi-
tion and also inhibition of prey-killing bite, which in part accounts
for the fact that they engage in sustained bouts of contactual and
play behaviors compared to coyotes and foxes (Fox, 1975). The
conclusion, then, is that there is a greater degree of genetic control
of bite intensity in domestic dogs (beagles), while in coyotes selec-
tive control is effected more through social and experiential influ-
ences.

In some breeds of domesticated dogs, notably in the bird dog
pointers and retrievers, there has been rigorous artificial selection
for "soft mouth," for the dog must not mark the game. The prey-
catching sequence in such breeds consists of orientation, approach
(or track), and either point or grab and carry (retrieve), the killing
bite and head-shake actions being selechvely eliminated.

Vauk (1953) reports that play with prey (and vrith inanimate
objects) develops at different ages in different breeds of dog. He
concludes that domestication has modified or completely inhibited
prey-killing, in certain breeds; some, like the pointer, develop
fixed, exaggerated signs. The effects of selective breeding and
training on prey-catching and killing behavior in the domestic dog
is summarized in Table Vn. It is evident that domestication and
selective breeding may lead to a breakdown in the normal temporal
sequence at one of several points by intensifying or inhibiting cer-
tain action patterns. Various breeds, selected for specific hunting
and tracking tasks, clearly demonstrate this subtle aspect of domes-
tication. The capacity of domestic dogs to hunt and kill prey and
live independently has been demonstrated in feral dog studies (see
Chapter 3). Therefore, these abilities are not eliminated through
domestication, but rather, motivation, threshold of different ac-
tions, sodalization, training, and dependence upon man are some
of the interrelated variables that must be considered in evaluating
the effects of domestication upon animal (and human) behavior.

Two 8-weck-old Australian dingos (undomesticated) were re-
cently tested. All lolled and ingested prey on the first test, with an

Effects of Domestication on Prey-Catching and Killing

131

are highly innervated in carnivores to provide tactile feedback so
that the positioning and pressure of the bite can be exactly con-
trolled.

Summary

In order to identify possible effects of domestication on behavior,
prey-catching and killing behavior was studied in domestic dogs
(beagles), coyotes, and Fj and F3 generation beagle X coyote hy-
brids. The full temporal sequence of prey-catching and killing, to-
gether with efficient dissection and ingestion of the prey evident in
coyotes at 8 weeks of age, were respectively truncated and disor-
ganized in the domestic dog and hybrids. These changes in be-
havioral organization and temporal sequencing of action patterns
are discussed in relation to differences in arousal (motivation) and
response threshold. Partial and complete inhibition of the killing
bite was found in hybrids and beagles, respectively, this being a
major factor in the truncation of the full prey-catching and killing
sequence. Domestication may lead to truncation of this temporal
sequence at one of many points by intensifying or inhibiting certain
action patterns through selective breeding.

130

THE dog: its domestication and behavior

not kill live prey showed playful and actively submissive actions
toward the prey, but after being forcibly fed dissected prey, they
essentially learned that it was food and from then on would attack
and kill prey and direct no further social behavior toward it.

Thus, some species (wolves) have to learn what to kill and
what is food, while in others (foxes and coyotes) learning has less
influence (Fox, 1969). The argument here is that there may be some
connection between intraspecific aggression, general bite inhibi-
tion, and prey-kilhng. By analogy also there may be considerable
resistance and conflict in training a dog to selectively attack certain
people (on command or without command in a certain place they
may ^ard) and not to attack other people or in other places, when
there is socialized attachment to people and generalized bite inhibi-
tion. Therefore, no single factor is wholly responsible for prey-
killmg and social inhibition of biting in wolves or of selective disinhi-
ition of attacking and biting man in trained guard dogs. Genes,
early experience, social attachment, and training all contribute to
^ expression of these motivationally separate activities of

I^ey-killing and agonistic behavior. Social attachment to prey
(Kuo, i960) may inhibit prey-killing behavior; intraspecific bite in-
associated with a more generalized bite inhibition
w ich has to be disinhibited in relation to prey in some species
sue as the wolf. Thus, although motivationally distinct, certain
vanables may affect both aggressive behavior and prey-killing reac-
tions equally or in a similar direction; hence perhaps the frequent
hatior'^”*^^^^ association between hunting and aggressive be-

Similarly, wolf cubs show more social inhibition of aggression
and ^ting conspecifics than coyotes and foxes who at an early age
may a conspecific reflexively as though its fur and movement
re eased a prey-killing response. Social experience in controlling
Dite inhibition in coyotes and foxes is therefore important in early
lire, while release from a more generalized inhibition of biting is
necessary for some wolf cubs to secure live prey.

In future studies it would be advantageous to use pressure
transducers to provide quantitative data on developmental
changes and on spedes and breed differences in bite intensity and
control. The contribution of genetic and experiential influences
might then be eluddaled. Ewer (1973) notes that the canine teeth

VII

Interspecies
Interaction Differences
in Play Actions in
Canids

Introduction

Most studies of social behavior have focused upon intraspecies
interactions. This study involves a little explored area of in-
terspecies social interactions, where the frequency of occurrence of
clearly identified discrete action patterns may be modified in rela-
tion to the degree of dyadic compatibility within the selected con-

text of playful interaction. . , ,

Yves Rouget (personal communication) raised a red fox (yulpes
vulpes) with a domesticated dog. The impressive, but unquantified,
consequence which he recorded on film was the high frequency of
face-oriented pawing manifest by the fox. Such an action is rare in
foxes but is a common action pattern in domesticated dogs. This
kind of observation opens the question of the role of expenence m
the development and reinforcement of species-typical achon pat-

An incompatibility of reciprocal actions has been reported by
Blauvelt (1964) between sheep and goats raised together, no effec-
tive fights were seen because the sheep tried to butt the goat and
the goat tried to jump on the sheep. Interestingly, whom the am

133

Interspecies Interaction Differences in Play Actions

135

lowed by side-to-side head flexions which move fluidly down the
body, followed by rolling over, "spinning," running around the
partner, "diving" or exaggerated approach or withdrawal (Fox,
1971b). From the play-soliciting bow, the head may be twisted up-
ward toward the forelegs, throat, or cheeks of the congener and
bite intention lunges or incomplete bites are then executed (Figure
i).

An action resembling the play-leap occurs when one coyote is
chasing another; the action is of Jow amplitude and is followed by a
bite or bite intention on the rump or shoulders of the partner or by
a forelimb clasp (Figure 1). This clasp, which is often never fully
executed, consists of seizing the partner around the waist and fre-
quently following with a scruff-oriented bite. The partner may

Figure 2. Although scruff-oriented biting in coyote (D) nffen oc-
curs during play-fighting, the coyote usually approaches low (A
and B) and makes an upward directed bite from the <C) play-
soliciting posture and rarely directs vertical leaps at its partner.

134

THE DOG: ITS DOMESTICATION AND BEHAVIOR

mal attacked was determined by early sodal experience, but the
manner in which it attacked remained unchanged.

A consistent feature of social play in C. familiaris, Fi C. familiaris
X C. lupus, and FjC. familiaris x C. latrans is the occurrence of the
play-leap (Figure i). This action may be preceded by a direct stare,
a play-soliciting bow or an incomplete leap intention which may be
repeated three or four times. The leap itself varies in amplitude and
may be repeated as the subject moves in on its congener. The latter
may also rear up and meet its partner with a reciprocal leap, and
both animals briefly stand and wrestle with the forelimbs and make
face-, cheek- and scruff-oriented bites or incomplete bites. Most
often the leap is followed by an incomplete bite or bite intention or
by a nose-stab which may be interpreted as a bite intention move-
ment. Occasionally this may be followed by a backward leap.

In the coyote, C. latrans, such sequences involving the play-
leap are not observed. In this species, a play-soliciting bow is fol-

Figure i. (A) Vertical leap of beagle during play which may be re-
ciprocated (B) and both rear up. Leap may be followed by face- or
cheek-oriented bite. (D) Playful standing cater by F, coyote x beagle;
<E) clasping by coyote during play-fighting.

Interspecies Interaction Differences in Play Actions

137

Table I.

SUMMARY OF PLAY-LEAP SCORES

Subject

With beagle

With coyote

With coydog

Beagle

8.5“ (2-16)'’

16.3 (1-37)

15.7 (2-35)

Coyote

1.0 (0-2)

2.5 (0-7)

0 (0)

Coydog

12.7 (1-31)

5.7 (4-9)

16.0 (7-22)

“Average scores per individual.
•"Range shown in parentheses.

coydog than with its own species and this may account for the
lower scores in beagle pairs. , , . ,

An intriguing finding was that the coydogs had high frequen-
cies of play-leaps when paired with a conspeafic or beagle but
lower frequencies when with the coyote. This was attributed not to
a lowered activity or motivation, for play was intense, ®

possibility that the hybrids had a more fleidble or adaphve play
repertoire which could be reciprocaUy matched with a beagle or
coyote partner. In contrast, the beagle when paire
showed the highest average frequencies of P ^7' were

not reciprocatel No significant differences attnbutable to sex were

evident in any of the three canid types.

Discussion

These data support the are bi'com

quences of beagle and coyote ^unng intentions of

The bow followed by upward direc play-attack; the play-

the coyote may be regarded as Im ^ P piay.atiack

leaps of the beagle may be attack orientation.

The coyote, in responding to j or hip-slam to block or

twists away and usually executes a K-agle but

deflect the attack; such actions were nrobablvcontnbuted to

were evident in the coyote x dogs and most prob .

136

THE dog; rrs domestication and behavior

attempt to avoid being clasped by executing a hip-slam or twisting
around to face its congener. When facing, rearing up may occur,
followed by face-oriented bites or bite intentions and pushing with
the forelimbs.

Materials and
Methods

With these species-characteristic actions in mind, it was decided to
evaluate the compatibility and reciprocity of such action sequences
in variously combined pairs of subjects, by recording the frequen-
cies of play-leaps. All dyads were of opposite sex, each approxi-
mately 12 months of age. Each pair (of noncage mates) was ob-
served for 10 min play in an 8 x 8 ft soundproofed arena equipped
with one-way observation windows. Two observers recorded the
frequency of play-leaps for each pair, dyad combinations consist-
ing of beagle X beagle, beagle x coyote, beagle X coydog, coyote X
coyote, coyote x coydog, and coydog X coydog. A total of 20 dyad
combinations were tested in 12 subjects (4 beagles, 4 coyotes, 4
coydogs, with 2 females and 2 males per species).

Results

The findings are summarized in Table I. It should be noted that
both coyote and coydog showed hip-slams during play, but the
frequency of this action was not recorded. Also, several of the low
amplitude play-leaps recorded in the coyote were difficult to dis-
tinguish from the incomplete action of rearing up to clasp the part-
ner with the forelimbs.

The high frequencies of play-leaps in pairs of beagles, pairs of
coydogs, and beagle X coydog reflect the reciprocal nature of this
action. In all dyads where one individual scored over 15 play-leaps,
then its partner likewise had a comparably high score.

The beagles were more active when paired with a coyote or

Irtferspecies Interaction Differences in Play Actions

139

Summary

Spedes-typical patterns were identified in wild and domesticated
canids during play in dyads of the same or different species. While
these actions are under relatively rigid genetic control (i.e., are
inherited) other dimensions of the jjlay sequence may be more eas-
ily modified contextually. Thus, not only species and context de-
termine the frequency of occurrence of certain actions but also the
compatibility of the interactee, be it of the same or of a different
species. The long-term consequences of interspecies interactions
under conditions of captivity may lead to more permanent changes
in the occurrence, frequency, and amplitude of certain species-
typical, ''fixed" action patterns. This may be an additional variable
to consider in investigating the complex influences of domestication
upon behavior and warrants further investigation.

138

THE DOG: ITS DOMESTICATION AND BEHAVIOR

their ability to maintain reciprocal synchrony with the coyotes dur-
ing play bouts. These hybrids also showed the play-leap action
which enabled them to sustain reciprocal play actions with the
beagles.

■pie role of early experience and of social factors should also be
considered, and this point is emphasized here rather than earlier in
this book, with reference to the prior experiences of our canids.
The criticism that might be leveled at this study to the effect that
the beagles had no prior opportunity to interact and modify their
achons with respect to coyotes (and vice versa) is untenable. All
beagles and coyotes used in this study had been reared together
from approMmately 8 weeks of age, although tests were conducted
using indmduals from different cages. It might be expected that
coyotes with such early social experience might acquire the play-
leap achon of the beagle. But this was not evident; thus supporting
the nohon that the play actions studied are species-characteristic
mav h/ Tfi frequency and temporal sequencing of which
may be modified contextuaUy as in certain dyadic interactions,
the interaction with a different species may modify

oi^nH of certain action patterns, more pro-

long-lashng effects on interspecies interactions under
evahiateH* ^®.“®‘“"®^®nforced contact in captivity remain to be

" "t’ '"‘i^nge in the structure
anH fr occurrence

? elicitation may be altered, especially within the

and dncr u ’^"^^J^^cation, as when different species (cat

and n d ; I’”ses, etc.) are raised

on tnrh H ® influence of human behavior

canid evaluated. One classic example is the

canid analogue of the human “gieeKng grin" (which is quite dis-

(Fox iQ'^W ™ssive grin and open mouth play-face of canids
onw'h®^^ ®’‘P''®“i°n is mimicked by the dog and has

anLorn^’''" interactions and not between dog

famar d^' " ^'®°.^''nience that the capacity to elaborate this

dPvtln^ iines of dogs will never

ITp wL 1 I “Pi*® others from the

dov I -i*^ -.vi parents have this facial expression will quickly

develop It either spontaneously or with a little human reinforce-
ment.

VIII

Socialization Patterns
in Hand-Reared Wild
And Domesticated
Canids

Introduction

This chapter offers a number of observaHons about expenences
with hand-raised and socialized canids, including wo ves, coj^ es,
golden jackals, and red foxes. When assembled, these ^necdo es
constitute a long-term experiment on these spec.es caP'.v. t ,
where their social reacHons toward peopl^an be contrasted ..
their natural or innate social tendencies. The
long to three basic social types (Fox, 1975); (covotc

(red fox), transitional Type II with

and golden jackal), and the social Type t mnsiclhtion

legiaLs and pack formation (wolO-

of primarily inLrited tendencies, interact.ng -'h expenen .a ac

tors during early life, is a ■^njordeterm.nan of sckuI Khm .or an^^^

of the capacity "to form --nta.n s-al r«
life. In this chapter, it will bo shown how the in
of social ->.-.>ionships in the^ thre^
reactions toward people aiter inej
dalized to human beings.

Socialization Patterns in Wild and Domesticated Canids

H3

patterns of human contact in no way resembled the aforemen-
tioned restricted rearing and socializing conditions.

The distinctions made by Scott (1968b) between primary and
secondary socialization should be emphasized here. Primary so-
cialization refers to those social relationships which are established
early in life during a critical period (Scott 196^) between a wolf
cub, its parents, and littermates. Secondary socialization refers to
those social relationships that develop at a later age, as between
other adults of the pack that breaks up during the denning season.
The possibility that species differ in their capacity to establish and
maintain secondary social relationships will be explored in this

chapter.

Primary social bonds may facilitate secondary socialization; a
domesticated dog, for example, if raised with human beings early
in life is able to generalize from these primary socializing experi-
ences and is friendly toward strangers at a later age. Some breeds,
notably the guard dog types, seem to have a more limited capacity
to form secondary social relaHonships and are very wary of stran-
gers, The relationship between this phenomenon and the compara-
ble behavior in the wolf will be discussed subsequently.

Woolpy and Ginsburg (1967). in their extensive sl^ies of wolf

socialization, have shown that adult wolves caug t vvi , wi car

ful handUng, wiU become socialized to their handler and on the
basis of such experience tend to generalize

strange people as well. These findings put a different light on the
critical period hypothesis but in no way refute it. ins .“^6 '''^P
that coyotes do not generalize or form 'miin

ships to the same extent as wolves, ""f 'V";;;;';,"

attached to the principal person responsible S P

socialization. This is the key to the present report;
extent do the socialization capacities and ‘

beings of captive canids match their natural
ecologically adaptive, genetically determine soci. p,

Results

For convenience, the results are broken "

tions, and each section is discussed separateir The se-ctu n.

142

THE dog: its domestication and behavior

Materials and
Methods

The following observations are based on experiences with is wild
canids, all hand-raised from 2-6 days of age and kept in captivity
under similar conditions of handling and management until sexual
maturity (2 years) and in some animals up to 4 years of age. Sub-
jects include i male and 2 female wolves, 2 male jackals, 2 female
and 1 male red fox, and 5 female and 2 male coyotes. In addition, 6
beagles, 8 pointers, and 4 Chihuahuas were raised under similar
conditions for comparative purposes.

theoretical premise

reactions of these animals, all with

anZnw^ Iiand-reared them,

and toward research assistants who later took care of them. Over a

son who almost daily contact with the per-

son who had raised them and also regular contacts with an average

siHriTr^ ‘w°/e‘r'ale assistants per week, who were respon-

was aoor f ^ ^ate of persomel

o^^ ^ “ months. The social conditions in

of human contact were therefore
studW h" " ‘l^at those students

S fn nontraumatic contact in-

viiums werr?®" P^r year. Occasional

Tn summf i“”'" 1 faciUty on a random basis,

ma r^xT ^nntact/Lposure to approxi-
mate whethiR^"°‘’f ^ach year. It is imposSble to

sociali^tnot '""‘act facilitates Lcondary

soda cotftm, the annual turnover eventuaUy leads to

H experience with such

nt^sr *at most dogs readily accept

one ort!? ■ ^ Coyotes raised exclusively by

ZZIT. ’’'Tk "“'=P‘ a ehange in personnel or

change of sex of handler and will show fear or aggression toward
the new person (personal observations). In the priient chapter the

Socialization Patterns in Wild and Domesticated Canids

H5

arbitrarily categorized as follows: permanence of primary social
attachments, capacity to form secondary or subsequent attach-
ments, changes with sexual maturity, socio-sexual discriminations.
In each section, the behavior of the canids toward human beings is
compared and contrasted with their intraspecific behavior in both
captivity and in the wild. (See also Table 1.)

PERMANENCE OF PRIMARY SOCIAL ATTACHMENTS

All subjects, irrespective of species and sex, remained attached to
the author who hand-reared them. They all showed active submis-
sive greeting and passive submission when physical contact was
made. The red fox, however, tended to avoid physical contact in
spite of showing active submissive greeting and occasional play
solicitation. When contact was forced, the fox would remain pas-
sive or attempt to escape and occasionally bite. The active greeting
by this species involved virtually no physical contact in contrast to
the face-oriented licking, pawing, and jumping up of the other
canid types. In the wolves, actual contact initiated by the anirnal,
including rubbing and contactual leaning, was of greater duration
than in the Type II canids (coyote and jackal).

It may be concluded that in these three canid types, proximity
tolerance was lowest in the red fox (Type I), while maintenance of
close proximity was greatest in the wolf (Type III). The coyotes an
jackals (Type II) were essentially intermediate; in the coyo'es.
greater individual variation in proximity tolerance devc ope wi
increasing age. Some individuals after 2 years of age i\ou on >
briefly greet, and then withdraw, while at an ear ler age uou
maintain contact and display active and passwe su mi^ion or a
longer period. Others did not show such a change '' ' _

offensive or defensive aggression W.1S ever disp a> e ) ,

and III canids toward the author, while such occurrences were not
infrequent in the Type I canid. In ‘'vo instances the author as
subjected to redirected aggression during a con ic '
coyotes; one of the pair redirected its attack on the hnneher. . u
bites being inhibited, hmvever. No "penences Im c ^emred
with the wolves. All domesticated hand-raise months of

tached through to maturity, some individuals after 5

Table 1.

SUMMARY OF SOCIALIZATION PATTERNS IN CANIDS

Socialization Patterns in Wild and Domesticated Canids

have hand-raised wolves. What evokes fear of men — odor (andro-
gen metabolites) or human male body langage— remains to be
evaluated. With sexual maturity (2 years) this wolf also showed
increasing aggression toward conspedfic females (and also female
domesticated dogs) but displayed active and passive submission
toward male wolves and large male dogs. These reactions were
more intense during estrus in February.

Of the two male jackals^ one was more timid and would only
greet and maintain proximity with the author; fear of strangers was
evident before 1 month of age in this animal. The other jackal
remained friendly toward strangers until approximately i year of
age. Increasing ambivalence between approaching, greeting, in-
vestigating, and solidting grooming and withdrawing was evident.
Subsequently, this animal began to threaten strangers in a highly
ambivalent defensive-offensive display, showing more offensive
aggression toward men than women. By 2 years of age, he ha
attacked and sUghtly injured two females who were regular atten-
dants and one strange female; strange males could not even enter
the cage.

The seven coyotes were similarly friendly toward strangers,
new, and regular personnel irrespective of sex until after 1 year o
age. After this age, aU animals began to show increasing hesitation
to approach, creet, and investigate strangers, with only one excep
Hon. This animal, a female, instead showed increasing aggression
toward males and occasionally toward females. One ma '

females of the remaining six coyotes became increasing y ‘
and would remain in their nest boxes when either regu ar pc
nel or strangers came close. The remaining I ree
indifferent, in that they would not greet strangers or gu pc

sonnel, but they did not show "ge provides an

Another male coyote, obtained at 2
additional example. This animal had been L

women and remained attached to them as f,._ni«and was

years of age he was kept and mistreated by a This covote

<,ve,l»,lly gl... ,0 Ih, mn-pl”"

146

THE dog: its domestication and behavior

age showing a tendency toward passive submission rather than
active submission (friendly greeting) when handled by the author.
This correlated with a tendency toward avoidance (fear) of strang-
ers in such dogs, but after brief exposure, unlike the wild canids,
they would become more active and begin to accept strangers. This
capacity for secondary socialization (or social potential per se) con-
trasts the more limited capacity evident in nondomesticated canids
(see below).

CAPACITY FOR SECONDARY OR SUBSEQUENT ATTACHMENTS

Between 1 and iVr years of age, the wolves began to show increas-
ing wariness of strangers. Prior to this time they would display
active and passive submission toward a stranger, but after 18
months of age, an increasing flight tendency was apparent, more
so in the male wolf of the pair. Between 2 and 3 years of age, this
male became increasingly wary of strangers, but if the person was
introduced by the author and remained passive in a squatting posi-
tion, the male wolf would eventually approach and investigate,
vasdllating between active and passive submission and flight. The
female at this age would usually approach a stranger after a brief
period of ambivalence to investigate, greet, and solicit petting. The
female was more exploratory even at 3 weeks of age, and at 4 years
of age showed more positive social responses and less fear toward
strangers than the male.

During the heat period at 2 and 3 years of age, this female
showed less passive submission and displayed aggression toward
one of the male golden jackals in the opposite cage particularly at
these times. After 2 years of age, the male wolf began to threaten
certain visitors and at 3 years attempted to attack one student. All
agonistic behavior was directed against male human beings, never
against women. The possibility that the male wolf is making a
sexual discrimination, and upon what basis (possibly olfactory),
warrants further study.

A third female wolf recently raised in the same way developed
an inaeasing fear of strange male humans after 5-6 months of age;
adult females and preadolescent children were readily accepted
even 6 years later. This pattern has been confirmed by others who

Socialization Patterns in Wild and Domesticated Canids

149

domestic dog, sexual maturity is attained as early as 6 months of
age, but there is a delay in appearance of territorial behavior and
aggression toward strangers (both conspecifics and people) until 12
to 18 months of age, and in some cases up to 2 years. This latter
behavior may be hormonally dependent since early castration or
ovariectomy may prevent such behavior from developing (Brun-
ner, 1968). In the dog therefore, domestication may have caused a
split in the timing of gonadal and central nervous system matura-
tion and integration, precocious sexual development being one
selected attribute in domestication.

SOCIO-SEXUAI DISCRIMINATIONS

The marked discrimination of human sexuality by some of ^e
Type n and ni canids only emerged with increasing sexual mata-
rity. As far as could be ascertained, male assistants did not ^^^at
the animals and treated them in a comparable way o /

dress was also often identical, both sexes having long air an
jeans! This sex discrimination was most marked in e m e
and the male golden jackal, overt aggression being ‘°-

ward strange males at increasing frequency a er 2 y . . j
The author was attacked by a male wolf dunng the ftirmng of
the “Wolf Men," a television (NBC) , , been

stranger to this wolf, which, together wit 1 s ( Iban its
hand-raised but had infrequent contact wit pe^
handler. Prior to releasing the two wolves from *eir cage for him
ing in a large enclosure, the female, -'ho was m heat, sobated
author through the cage with full vagina) disp ay repeatedly

The male wolf intercepted her,

threatened the author. When released, xhis example

tacked the author (for further details see

Illustrates clearly one of the problems o ^eact toward

a consequence of socialization, they wo

human beings in a reviewed m exlat$o by

normally react to conspecihcs (this ^ ^ ,be attack was

Fox, 1968c, 1971b); the most logical the

that the author represented a s^xua pences of socializa-

attentions of the female. More exlrem

148

THE dog: its domestication and behavior

gression, in contrast to the purely offensive display toward
females.

Experiments moving coyotes from one enclosure to another
and rearranging pairs has revealed a tendency for females to be
more aggressive toward strange male and female conspecifics than
their male partners. A strange adult male, dominant to the resident
female but subordinate to the resident male 6 months after intro-
duction, caused a dramatic change in the pair-bond. The female
developed an allegiance with the strange male, who became domi-
nant over her original partner, but in her absence remained sub-
ordinate. Any female coyote placed in the cage with these animals
would only be attacked by this resident female. The possibility that
the dyadic relationship in the coyote is matiifocal (in that the
female is the controller and selector of the partner and regulates
intraspedfic proximity) deserves further analysis in the field.

In the red fox, wariness of strangers was consistently seen as
early as 4 months of age, and in only one instance was active
submissive greeting displayed toward a person other than that
who had been responsible for hand-rearing. This lack of generali-
zation clearly demonstrates the limited capacity of the red fox to
develop secondary social relationships.

It should be emphasized that all subjects had more or less the
same exposure to people after weaning had been completed be-
tween 3 to 4 weeks of age.

CHANGES WITH SEXUAL MATURITY

With the exception of the red fox, many of the above changes in
behavior correlate with the attainment of sexual maturity. The
Type II and III canids show a gradually increasing tendency to
avoid strangers after 4 to 5 months of age, which is clearly evident
at 12 to 18 months and tends to persist. In captivity, the female
coyotes have shown estrus at 10 months of age, although they do
not normally show estrus or breed until the second year of life. The
female wolf had her first heat at 2 years, the male showing little
interest until the following year. The later emergence of defensive
and offensive aggression rather than avoidance reactions toward
strangers may be related to this late sexual maturation. In the

Socialization Patterns in Wild and Domesticated Canids

mally leads to dispersion around 5 months of age in the Type I
canids and around 10 months in the Type II canids (Fox, 1975). In
the more sociable Type III canid, the care-dependency relationship
gives way to a dominance-subordinate relationship combined with
allegiance and affection (Fox, 1972). The capacity to form secondary
social relationships and to form a stable social hierarchy is limited
in the Type II canid and more or less absent in the Type I solitary
canid.

Compared to the domestic dog, the wolf seems to have a lesser
capacity to form secondary social relationships. There may be a
period after which such relationships are difficult to establish in the
wolf. This may be correlated with the fact that the wolf pack, espe-
cially the upper echelon of mature dominant wolves, is closed to
strange wolves. Strangers are avoided or driven away, while lower
ranking yearling wolves will more readily accept strangers and
may even leave the main pack (Fox, 1973)' social behavior of
the more mature wolves is an ecologically adaptive sodal
mechanism that regulates pack size. In the domestic dog, this e-
havior may have been selectively eliminated in many breeds so that
as adults they will readily accept strangers. The possibility of infan-
tiUsm (Zimen, 1970) or neotenic perseverance of infantile care-
solidting and submissive behavior in the domestic dog may aci
tate the generalization of secondary sodaiization as well as inaeas-
ing proximity tolerance. Most significant, however, is * ^
for innate sodal predispositions which may fadUtate t e ’

cation of some spedes but be a major limiting factor in o ®

different temperament and capadty to develop secon ary so
relationships.

Summary

The sodal consequences of hand-rearing have

their actions to their foster parent and to ot er ■ socio-

been reviewed. The canids studied ^ .md

ecological types, namely J D^fferencfs in permanence

intermediate coyote and golden sccondarv' social

of primary sodal attachments, capaatj t

150

THE dog: its domestication and behavior

tion are well illustrated in various spedes of birds (Klingbammer,
1967) and zoo animals (Hediger, 1950) which, when hand-raised,
later show a marked sexual preference for human beings. The con-
sequences of sodalization with man are less dear-cut in the various
canid spedes. All types studied remained attached to the person
who hand-raised them up to their current age of 4 years.* The
solitary Type I canid shows a limited capadty for generalization or
for the formation of secondary social relationships. This correlates
well with the socio-ecology of this Type (Fox, 1975). Several indi-
viduals of Type n showed a lesser capadty to form secondary social
relationships compared with the wolf. Recent observations of
dingos reveal a similar Type II pattern of sociability, a finding ver-
ified by Corbett and Newsome’s (1975) field study of their social
ecology. There is evidence that the Cape hunting dog, another
Type HI canid, also has a considerable capadty to form secondary
social relationships with people after being hand-raised. People
who have hand-raised Mexican wolves report that their sodaliza-
tion patterns resemble the coyotes more than the typical wolf. This
accords well with the social ecology of this subspedes (Cams lupus
baileyi) which is closer to the coyote in social behavior, rarely form-
ing packs and more commonly being seen alone or in pairs.

Conclusions

The main conclusions of this study are summarized in Table I. In-a
review of sodalization problems in various nondomesticated,
hand-raised carnivores such as ocelots and raccoons (FoX/ 1972)/ a
common finding is increasing aggression and unpredictability of
behavior toward strangers, and increasing intolerance when disd-
plined by the owner. The change from the care-dependency rela-
tionship of the primary sodal attachment to a more independent
relationship is typical of the relatively solitary carnivores
(exemplified by the red fox in this study). Increasing proximity
intolerance and parent-infant and infant-infant aggression nor-

•Incrcasod aggression and assertion of rank, especially during the breeding season
may not be evident in the wolf until 4 to 5 years of age in both males and females.

IX

Stages and Periods in
Development:
Environmental
Influences and
Domestication

Introduction

The purpose of this chapter is to review a number
serve to illustrate various principles of neural and e
velopment. An attempt is made to bring together severa i
topics from the ethological and psychological literature
formulate some general concepts pertaining to t ® .

interpretation of critical and sensitive penods m e P
These periods represent developmental stages a w inne-

mental (i.e., human) influences can have .'“"t

lasting effects upon later behavior, phys'°'°^' to describe

tivity. A number of terms in current use that P . ■

some of these developmental phenomena are ^

to clarify the use of such terminology in this someiv ‘ ^1

of developmental psychobiology. Structura an . ^,5 of

velopment of the ca^ne brain is correlated -'h
behavioral ontogeny underlying the critical pen

■5J

152

THE dog: its domestication and behavior

relationships, changes with sexual maturity and socio-sexual dis-
criminations are discussed. The various relationships between so-
cialized canid and man reflect the innate social capacities of the
species in question and also correlate with the socio-ecological pat-
terns of the species under natural conditions. In other words, it is
shown that the socio-ecologically adapted temperament and social
capacities of each canid type are primary determinants of their
reactions toward human beings after early socialization with man.
Where relevant, the social behavior of domesticated dogs is com-
pared and contrasted with that of the wild species; in contrast to
the dingo and wolf (proposed ancestors of C. familiaris), most
breeds of dog have a greater social potential or ability to establish
secondary social relationships which, in contrast to these other
species, must be a consequence of domestication per se.

Stages and Periods in Development

155

Also, the size and elaboration of the processes of fibrous astrocytes
closely resemble the adult between 5 and 6 weeks of age (Fox,
1971b).

From these observations, it is apparent that neocortical
neuronal maturation precedes myelinization and that the second
myelinogenetic cycle does not begin until neuronal and glial de-
velopment is well advanced.

Detailed quaBtative studies on the postnatal development of
the canine EEG during wakefulness, quiet sleep, and paradoxical
(REM or activated) sleep show that the first signs of slow wave
(possibly thalamocortical) activity appears around 17 days, and by 4
weeks of age, a clear distinction between various states of wakeful-
ness and sleep can be detected on the EEG (Fox, 1971b). Relatively
mature patterns of electrocortical activity are present at 5 weeks of
age, at which time the percentage of various states of sleep and
wakefulness are relatively mature. Only slight qualitative changes
in EEG occur after 5 weeks of age, notably the preponderance of
fast spindle activity at the onset of quiet sleep, a gradual reduction
in amplitude and increase in fast frequency components during
wakefulness, and a more gradual increase in amplitude during
drowsiness and quiet sleep. In close temporal association with EEG
development, auditory and visual evoked polenHals, in terms of
latency, attain mature characteristics between 5 and 6 weeks of age.

Behavior Development in the
Dog

Several parameters of behavioral development in this species have
been investigated and correlate well with the above data on ChJS
development. The dog is neurologically mature at 4 weeks of age
with the exception of equilibration and adult locomotor abilities
such as running and leaping (Fox, i9;nb). Between 3 and 4 weeks of
age the pup begins to interact with its socioenvironmental milieu
and to develop primary social relationships or emotional attach-
ments. This marks the onset of the critical period of socialization

154

THE dog: its domestication and behavior

The degree of plasticity and adaptability of the developing
brain and behavior can be evaluated, for example, by environmen-
tal manipulations such as sensory deprivation, social isolation, or
excessive stimulation (handling, experiential enrichment) at dif-
ferent ages or stages of development. These various treatments that
may modify brain and behavior development are considered in
relation to animal domestication.

Structural and

Functional Development of the
Canine Brain

Myelinization of the canine central nervous system (CNS) occurs in
two cycles. At birth and during the first 3 weeks, there is a gradual
increase in myelin content of the spinal cord occurring earliest in
the cervical region and in motor roots prior to sensory roots. The
corticospinal tract is the last major efferent pathway to develop.
Subcortical structures posterior to the posterior commissure also
begin to myelinate during this early postnatal period (Fox, 1971a).
Cranial nerves associated with feeding and cephalic cutaneous
sensitivity (facial and trigeminal nerves) and with balance and
body-righting (nonacoustic portion of the 8th cranial nerve) are
well myelinated at birth. Between 3 and 4 weeks of age, the second
cycle of myelinization occurs, anterior to the posterior commissure;
specific and nonspecific thalamocortical afferent fibers become
myelinated (Fox, 1971a). The various regions of the neocortex do not
show an equal increase in myelinization. Myelin is first detected in
the somatosensory area at 4 weeks and by 6 weeks of age is more
evenly distributed in other regions such as the visual and auditory
cortex. The frontal lobe shows the most gradual myelinization.

In contrast to this more gradual myelinzation of the neocortex,
neuronal development in terms of cell density, neuronal size, and
apical and basilar dendritic complexity, attains relative maturity in
the sensorimotor, visual, and auditory cortex by 6 weeks of age.

Stages and Periods in Development

^57

with the mother but with a human being, a flashing red light, or a
moving cardboard box, they become preferentially attached to the
species or the object with which they have been raised. Some re-
versal is possible during subsequent weeks, but there is strong
evidence that this early exposure results in a very specific and
enduring attachment — so enduring that at maturity social and sex-
ual behavior may be directed toward the same stimulus to which
the bird was imprinted early in life.

Lorenz (1970) hand-raised jackdaws and crows and found that
when they reached maturity they would show courtship behavior
toward him and would attempt to mount his hand. During the
time when they would normally be taking care of their own young,
they would attempt to stuff grubs into his ears. Hediger (1950)
describes the experience of one of his zoo keepers who hand-raised
a male moose. When the moose reached sexual maturity, the
keeper led it into a field of female moose, and the young bull
became sexually aroused. Instead of directing his sexual behavior
toward a female moose, however, he attempted to mount his
keeper. Such bizarre behaviors are good examples of the enduring
effect of imprinting in determining later social and sexual prefer-
ences in various animals. Klinghammer (1967) has shown that the
effects of hand-rearing in various species of pigeons can vary. One
species, for example, if hand-raised, shows a sexual preference at
maturity exclusively for its human handler rather than for its own
species. Other closely related species show a reversal at maturity;
although still friendly toward the human foster parent, they are
only sexually attracted to their own species. Klinghammer also
identifies a third category of pigeons in which individuals have the
best of both worlds and show sexual behavior toward both their
own species and the foster parent!

It is also known that the hvo sexes of a given species arc not
affected in the same way by imprinting. In mallards, for example,
Schutz (1965) finds that early imprinting in the males later deter-
mines their sexual preference while female mallards, regardless of
imprinting, show an innate preference for males of their own
species. This was confirmed by raising male and female mallards
with different species of ducks. When they reached matunty, the
male mallards preferred the species with which they had tx'en

156

THE dog: its domestication and behavior

(Scott and Fuller, 1965), and the neurological and behavioral events
are briefly reviewed since they are important factors which under-
lie the beginning of the critical period.

The data on the developing canine brain serve to demonstrate
the temporal coincidence of development and maturation of sev-
eral interrelated structural and functional parameters. This coinci-
dence, which occurs between the fourth and fifth postnatal week in
the dog, may be termed a period of integration. It is at this time
that the several parts of the developing nervous system show both
structural and functional integration, which marks the beginning
of a relatively mature organizational level of activity.

At this time, the organism begins to interact rather than react
with conspecifics and through social experiences with both parent
and peers develops emotional attachments to its own kind or to
man. If denied human contact during this critical period from 4-12
weeks of age, it will subsequently avoid human contact (Scott and
Fuller, 1965). Such dogs are fearful of humans and are virtually
untrainable. The fear period which develops after 8 weeks of age
limits the capacity to develop new social attachments and essen-
tially terminates this critical socialization period. Thus even in a
domesticated species, lack of exposure to man during this forma-
tive period (when brain centers are integrating and emotional reac-
tions developing) will greatly limit the social potentials of the
species.

Imprinting and
Socialization

The concept of a critical period implies that experience at a particu-
lar time of development is essential for normal development to
continue. This is exemplified by the phenomenon of imprinting in
birds. Imprinting means attachment, and birds (such as ducklings)
that are relatively mature when hatched normally become im-
printed onto the mother during the first few hours after hatching.
If they are taken as soon as they are hatched and are raised not

Stages and Periods in Development

159

used three groups. One group was hand-raised exclusively with
people. The second group had contact only with other monkeys.
The third group had more or less equal contact with humans and
their own Idnd. The monkeys segregated into like groups lor social
play during early life and later clearly demonstrated sexual prefer-
ences based upon their rearing experience.

If social experiences are denied during the early critical period,
as demonstrated in our cat-dog socialization study, there is often
impairment in the subsequent development of social relationships.
Scott and Fuller (1965) showed that if dogs are denied human con-
tact until approximately 14 weeks of age, they are wild and unap-
proachable. Scott and Fuller demonstrated clearly the critical na-
ture of the socialization process and found that human contact
between 6 and 8 weeks of age seems to be optimal for the socializa-
tion of dogs (see Figure 1).

Although there is an optimal period for socializing pups, there
is evidence that dogs may subsequently regress or become feral.
The social bond with man may be broken when well-socialized
pups are placed in kennels at 3 or 4 months of age; by 6 or 8 months
they are shy of strangers and often of their caretakers if they have
not been handled much. In addition, they may be extremely fearful
when removed from their usual quarters. Their fearfulness is the
result of a combination of institutionalization and desodalization.

In connection with the phenomenon of desodalization, Woolpy
(1968) found that captive wild adult wolves can be sodalized in 6
months of careful handling and that when they are subsequently
given less human contact they do not regress or become deso-
dalized. In contrast, wolves that are sodalized early in life are like
dogs in that they will regress if they are subsequently given less
human contact. These findings suggest, therefore, that although
there is an optimal period early in life when sodalization can be
rapidly established, subsequent reinforcement is necessary be-
cause of some intrinsic instability of retention in young animals.
Woolpy (1968) concluded that

an important aspect of sodalization is learning to cope \dth a
previously unfamiliar enviroiunental situation in the presence
of unreduced subjective fear. . . . We have inteq^reted the re-
sults of both the tranquilizer and the sodalization experiments

158

THE dog: its domestication and behavior

cross-fostered while the female mallards tended to reverse their
social preference in sexual encounters and to seek out males of
their own species.

Dogs have been raised with cats during the critical period of
socialisation to evaluate further the effects of cross-fostering. In
this study a 3-week-old Chihuahua was placed with a litter of
4-week-old kittens. Five replications of the study were done (Fox,
1971a), and in each replication a battery of tests was given to the pup
at 12 weeks of age. It was found that the pups raised with kittens
made no social responses to their own mirror images; they literally
lacked species recognition. They also preferred the company of cats
to that of the littermate Chihuahuas which were used as controls.
The cats that had been raised with a dog were also sociable toward
dogs that had not been raised with cats. In contrast, cats that had
had no earlier exposure to dogs avoided contact with them — with
one exception. Salzen and Cornell (1968) did a comparable study
for color preferences in chicks, which is reminscent of the role
integrated schools might play in achieving interracial socialization.
They raised a green-dyed chicken with a group of red-dyed chick-
ens. When they placed these chickens together with a new group
of chickens that were all dyed green, the green chicken ran toward
his red companions and ^d not mix with the green group. Var-
iations on this experiment seem to confirm that allegiances and
social preferences are based on early social learning.

Even more subtle consequences of early rearing can affect so-
cial preferences. One experiment used three groups of pups. The
pups in one group were hand-raised and were exposed only to
humans. The pups in the second group were weaned early and
had almost equal contact with humans and dogs up to the point of
testing. The pups in the third group were raised with eadi other
(Fox, 1971a). When the pups were placed in new social groups, it
was found that they tended to segregate themselves in accordance
with their rearing history. The pups that had been hand-raised and
had no social experience with their own kind tended to stay together
while the pups that were weaned early and the controls tended to
segregate into like groups. Other aspects of social behavior were
affected also.

Sackett et al. (1965) found a very similar social consequence of
early rearing experiences in young rhesus monkeys. They too

Stages and Periods in Development

159

used three groups. One group was hand-raised exclusively with
people. The second group had contact only with other monkeys.
The third group had more or less equal contact with humans and
their own kind. The monkeys segregated into like groups for social
play during early life and later clearly demonstrated sexual prefer-
ences based upon their rearing experience.

If social experiences are denied during the early critical period,
as demonstrated in our cat-dog socialization study, there is often
impairment in the subsequent development of social relationships.
Scott and Fuller (1965) showed that if dogs are denied human con-
tact until approximately 14 weeks of age, they are wild and unap-
proachable. Scott and Fuller demonstrated clearly the critical na-
ture of the socialization process and found that human contact
between 6 and 8 weeks of age seems to be optimal for the socializa-
tion of dogs (see Figure i).

Although there is an optimal period for socializing pups, there
is evidence that dogs may subsequently regress or become feral.
The social bond with man may be broken when well-socialized
pups are placed in kennels at 3 or 4 months of age; by 6 or 8 months
they are shy of strangers and often of their caretakers if they have
not been handled much. In addition, they may be extremely fearful
when removed from their usual quarters. Their fearfulness is the
result of a combination of institutionalization and desocialization.

In connection with the phenomenon of desocialization, Woolpy
(1968) found that captive wild adult wolves can be socialized in 6
months of careful handling and that when they are subsequently
given less human contact they do not regress or become dcso-
cialized. In contrast, wolves that are socialized early in life arc like
dogs in that they will regress if they arc subsequently
human contact. These findings suggest, therefore, that although
there is an optimal period early in life when socialization
rapidly established, subsequent reinforcement is nccessaiy be-
cause of some intrinsic instability of retention in young animals.
Woolpy (1968) concluded that

an important aspect of socialization is learning to cope with a
previously unfamiliar environmental situation in the presence
of unreduced subjective fear. . . . We have inte^reted the re-
sults of both the tranquilizer and the sooaliz.alion experiments

response score ?

3 5 7 9

WEEKS OF AGE

Figure i. Socialization in the dog involves an initial attraction
phase, but if denied human contact until first tested at 5, 7, or 9
weeks of age, fear and the tendency to flee increases with increasing
age. The optimal penod for socialization is between 6 and 8 weeks.
(Adapted from Scott and fuller, 1965.)

to indicate that the fear of the unfamiliar is the primary obsta-
cle to wolf socialization and that, whUe the overt responses of
fear appear very early in life, its subjective components con-
tinue to develop throughout at least the first year. Socializa-
tion must be conditioned in the presence of the fully de-
veloped subjective components of fear, and hence it cannot be
permanently maintained in juveniles if they are left to develop
fear responses subsequent to having become socialized in early
life.

There seems to be a comparable critical period for the de-
velopment of emotional attachments in children. Bowlby (1971)/

Stages and Periods in Development i6i

from the Tavistock Clinic in London, has placed much emphasis
upon the critical nature of infant socialization in determining later
socially adjustive behavior and in preventing delinquency and
even antisocial and sociopathic behaviors. It seems that in both
animals and humans the initial or primary socialization early in life
is the basis for the development of subsequent secondary social
relationships; if primary attachments are not made or are in some
way modified, the consequences for the later social adjustment of
the individual can be quite serious, and in terms of domestication
greatly limits the potentials of the animal for human use.

Socialization and
Overattachment

The ''perpetual puppy" syndrome may develop in adult dogs if a
symbiotic relationship has been maintained by overindulgent and
permissive owners, and severe anaclitic depression may follow
separation of the dogs from their owners for surgery, boarding, or
quarantine (Fox, i968d). It should be emphasized that as a conse-
quence of the Symbiotic relationship with the owner the pet may
develop a variety of care-soliciting (et-epimeletic) reactions such as
whining, jumping up, following constantly throughout the house,
crying when left ^one, and submissive urination. These symptoms
resemble regression in man to more neotenic or infantile behaviors.
Punishment after the disturbed pet (or child) has urinated, defe-
cated, or generally pestered the owner sufficiently may lead to a
masochistic form of reinforcement.

‘"Sympathy" lameness, hysterical paraplegia, and coxalgia
have been described in dogs (Fox, i968d) and are well documented
in man as attention-seeking reactions. Chess (1969) stresses the
correlation between acute illness and dependency behavior in
human infants. After recovery, the child attempts to maintain the
interpersonal relationship that brought him special attention dur-
ing the illness. Some overindulged pets have been kno^vn to refuse
to use one limb after surgery for congenital patella luxation because
they received so much attention and petting from (heir owners

i 62

THE dog: its domestication and behavior

while they were recovering. In extreme cases, muscle atrophy de-
veloped; in others, surgical recovery was complete, but the sub-
jects would suddently become lame when they were in an anxiety
or conflict-provoking situation and were soliciting the attention of
their owners.

Environmental and
Experiential Influences

Environmental and experiential factors which influence develop-
ment are reviewed briefly to provide a basis for further discussion
of the importance of environmental influences in the process of
animal domestication.

Schneirla (1965) has developed a very important theory involv-
ing the significance of approach-withdrawal processes in the or-
ganization and development of behavior. For instance, approach
and subsequent reward (such as contact comfort) are tied in with
parasympathetic arousal, while withdrawal from painful stimula-
tion is associated with adrenal-sympathetic arousal. Contact com-
fort, especially that associated with nursing or petting an animal,
causes parasympathetic arousal (exemplified in the human infant
by salivation, increased peristalsis, secretion of digestive juices,
general relaxation, and occasionally penile erection). Young ani=
mals derive considerable reward from contact comfort, grooming,
and nursing; if the same circumstances cause parasympathetic
arousal, then contact comfort, grooming, and nursing would tend
to improve digestion and assimilation of food as well as to facilitate
emotional attachment (Fox, 1975b). This "gentling" or petting phe-
nomenon is part of the taming-domestication process. Even gen-
tling a pregnant animal can result in offspring that are more docile,
and as emphasized by Denenberg and Whimbey (1963), this may
be a significant fact in the domestication process.

Several years ago Spitz (1949) demonstrated that inadequate
mothering, which may be reinterpreted as inadequate parasym-
pathetic arousal, led to a wasting disease in many orphan children.

Stages and Periods in Development

163

These children did not gain weight, did not adequately digest and
assimilate their food, and many succumbed to infections. When
Spitz initiated a regime of mothering, the infants began to gain
weight, and the rate of mortality decreased significantly. This ef-
fect, distinct from the handling effect, has been termed gentling.
Bleicher (personal communication) has found similar effects
in handled and nonhandled orphan pups raised in social isola-
tion.* No hormone has yet been identified, but it is possible
that the gentling of pregnant rats influences the parasympathetic
nervous system and that certain neurohormones affect the develop-
ing fetus.

A number of independent studies on rodents supports
Schneirla's theory, demonstrating that experimental manipula-
tions early in life can have long-lasting effects on later behavior.
For instance, if a 5-day-old rat is taken out of its nest and is placed
in a metal pot at room temperature for 3 min a day for 5 days and
then is allowed to mature, as an adult it is less emotional than
littermates not treated in this way. It may also be more resistent to
physical stress such as terminal starvation, cold exposure, and cer-
tain pathogens. This effect, which has been called the early han-
dling effect, seems to influence the adrenal-pituitary axis of the
developing rodent. Denenberg (1967), Levine and Mullins (1966),
and other workers have studied this phenomenon in great detail, ft
would appear that the stress to which the neonate is exposed in
some way affects the way in which it responds to psychological
and physical stresses later in life. It has been proposed that a hor-
monostat exists in the neurohypophysis which is "tuned" to the
adrenal glands during the sensitive period when the rat is between
5 and 10 days of age. When this is tuned experimentally by a
sudden elevation of corticosteroids during the sensitive period, the
hormonostat operates differently when the organism is stressed in
maturity. Typically, as Levine has demonstrated, the stress re-
sponse in adult control rats is somewhat maladaptive. There is a
long latency period before the neuroendocrine system responds,
but the eventual response can be long-lasting and at times may

These effects have also been found in calves (Professor C. Schivabc, personal
communication). Mortality rates were lower on farms where calves were handled
more ofter> and with more care and affeebon.

164

THE dog: its domestication and behavior

trigger the onset of Selye's stress syndrome. Rats handled early
have a much shorter latency of response, and the duration of re-
sponse is shorter. In effect, their reactions are not unduly delayed
and they do not overreact.

In some strains of mice that develop spontaneous leukemia,
the onset of the condition can be delayed by this handling proce-
dure. Handled mice also have a greater resistance than nonhan-
dled controls to Implanted tumors. There are also genetic or strain
differences, as demonstrated by Ginsburg (1968). For one strain, a
given level of stimulation might be excessive; in another strain, the
sensitive period might lie between postnatal days 6 and 10 or be-
tween postnatal days 10 and 15. Ginsburg emphasizes that in a
typically heterogeneous population, as in Homo sapiens, there
would be a normal distribution curve and that individuals would
have different sensitive periods as well as different response
thresholds. In view of these individual and genetic variations,
therefore, we must be very careful in making generalizations about
the handling effect.

Levine points out that in the relatively "overswaddled" labora-
tory environment neonate rats may well be understressed and
that the handling procedure is much closer to the kind of experi-
ence they would normally have in the wild. Under natural condi-
tions, the mother frequently leaves the nest to forage for food, and
there are various environmental changes, all of which could have
an additive effect resulting in an adult animal that is psychophysio-
logically superior to the "overswaddled" laboratory specimen.

In a study of the handling phenomenon in dogs (Fox, 1971a),
pups were subjected to varied stimulation — exposure to cold, ves-
tibular stimulation on a tilting board, exposure to flashing lights,
and auditory stimulation — from birth until 5 weeks of age. The
pups in the study differed from the controls in a number of ways
including earlier maturation of EEG, lowered emotionality which
enhanced problem-solving ability in novel situations, and domi-
nance over controls in competitive situations. Analysis of their ad-
renal glands indicated a fivefold increase in norepinephrine, and
studies of their heart rates indicated that a greater sympathetic tone
was developed as a consequence of the early handling stress. Meier
(1961) demonstrated comparable maturation of EEG and behavioral
changes in superstimulatcd Siamese kittens. (See Chapter 10.)

Stages and Periods in Development

165

A generalization cannot be made about optimal handling for a
given species. For instance, it appears possible that a given level of
early stimulation may produce psychophysiological superiority in
some human infants and pathophysiological inferiority in certain
other individuals. As Thomas et al. (1970) have demonstrated in
their longitudinal studies of human infants, handling has to be
very carefully adjusted to the basic temperament and autonomic
tuning of the individual. The effects of early handling appear to be
on the adrenal-sympathetic system, and the research evidence
suggests that early handling not only resets the pituitary-adrenal
axis but in some way influences autonomic tuning and tempera-
ment or emotionality.

To what extent current animal husbandry practices and vari-
ous handling regimes adopted in early domestication affect farm
livestock and house pets alike, remains to be evaluated. Those
breeders who have adopted the above handling stress program for
puppies and also the U.S. Army Veterinary Corps (Biosensor Re-
search Division) report extremely promising results in terms of
later stress — resistance, emotional stability, and improved learning
ability. This phenomenon may represent a new tool in animal hus-
bandry and domestication.

Prenatal Influences

Prenatal anxiety can have a significant effect on the emotionality
and the later personality developed of an infant (JoHq, 1969).
Thompson (1937) was one of the first to demonstrate the effects of
prenatal anxiety on the behavior of offspring. He developed a con-
ditioned emotional reaction in rats placed in a shuttle box;
whenever a bell rang, the rats had to jump to the safe side of the
box in order to avoid shock. The basis for an emotional reaction
was established by placing a barrier in the center of the box so that
the animals could not jump to the safe side when the bell rang.
Since the rats were never given shock when the bell rang after the
barrier was erected, the reaction was purely emotional. The off-
spring of pregnant rats that were stressed in this way were cross-
fostered by normal mothers in order to control for any postnatal

i66

THE dog: its domestication AND BEHAVIOR

transfer effects from the mother. Thompson found that the prenat-
ally stressed rats were much more inferior because heightened
emotionality interfered with their performance. He was able to
produce a similar effect by injecting adrenalin and ACTH into preg-
nant females.

Joffe (1969) has reviewed many such experiments including his
own on the Maudsley reactive and nonreactive strains of rats. He
was able to demonstrate genetic differences in susceptibility to
prenatal stress as well as differences correlated with the sex of the
offspring and the period during pregnancy when the stress was
administered. He discovered that premating stress could also have
a significant effect on the behavior and the reactivity of the off-
spring. These experiments emphasize that the developing
phenotype is modifiable not only after birth but also prenatally, the
effect being mediated by the neuroendocrine system.

These prenatal and postnatal studies in modifying the
phenotype also show that relatively little is known about how to
provide environmental stimulation and rearing programs which
will ensure that the developing phenotype can be optiiruzed or at
least how it can be protected from deleterious influences in food-
producing farm animals. There is much discussion about genetic
engineering and programmed breeding through artificial insemina-
tion, but the effects of environmental phenomena that are present
during certain critical periods in development have barely begun to
be explored.

Environmental Enrichment and
Deprivation

In The Descent of Man, Darwin observed that the domesticated rab-
bit has a much smaller brain than its wild counterpart and attrib-
uted this to the cumulative effect of generations of captivity during
which each generation received far less stimulation than it would
have in the wild. To test the effect of environmental stimulation on
captive animals, Bennett and his co-workers (1964) at the Lawrence

Stages and Periods in Development

167

Radiation Laboratory in California raised rats in enriched environ-
ments. Typical laboratory rats were placed in a large cage with cage
mates and with various objects (wheels, runways, and so on) to
manipulate. Under these conditions of environmental complexify,
Bennett found that the rats in the study became more exploratory
than the controls and that they developed significant differences in
brain size, in the depth of the visual cortex, and also in the turn-
over rates of brain acetylcholinesterase. More recently Bennett
demonstrated that the enrichment effect is dependent upon an
interaction between the inanimate environment and conspecifics.
That is, if a rat is raised alone in an enriched environment, pro-
found changes in brain and behavior do not occur; nor do they
occur if a rat is raised with companions in a barren cage.

Dogs have been raised under a schedule of paced increments
of experience, periodically taking some of them out of their home
cages and letting them explore an arena containing novel stimuli
(Fox, igyia). Those that were allowed in the arena at 5, 8, 12, and 16
weeks of age for a mere V2 hr per exposure explored increasingly as
they grew older and developed a preference for more complex
stimuli as they matured. Litfermates that were placed in the arena
for the first time at 12 or 16 weeks of age did not explore; they
withdrew or did not leave the start chamber, and many of them
were catatonic with fear. We are dealing here with an in-
stitutionalization syndrome; those dogs that did not have an op-
portunity to leave their home cages until sometime after 8 weeks of
age could not tolerate the complexity of the environment and so
they withdrew to avoid overstimulation. (Later some of the neuro-
logical effects of the emergence from isolation are discussed.) Sac-
kett (1968) adds support to our canid study in an experiment that
he did with rhesus monkeys. He found that as rhesus monkeys get
older, they prefer to look at picture cards of increasing complexity.

If monkeys are raised in isolation, at 6 months of age they prefer to
look at cards of less complexity than those which rhesus monkeys
normally prefer at that age. He concludes from these experiments
that in the absence of paced increments of experience the organism
will seek a lower level of stimulation and environmental complexity.

From these studies an arousal-maintenance model of
perceptual-motor homeostasis can be formulated wherein the op-
timal arousal or tolerance level is set early in life as a result of the

i68

THE dog: its domestication and behavior

quality and quantity of early experiences. The level is set low in
those animals that have had few increments of experience. If the
environment does not provide varied stimulation, the subject may
compensate by creating its own varied input by elaborating
stereotyped motor acts or by directing specific activities toward
inappropriate objects (such as copulating with its food bowl). The
stereotyped motor acts (thumb-sucking, self-clutching, and rock-
ing in primates) developed while in isolation may be performed
when the subject is in a novel environment and may serve to re-
duce arousal or anxiety because they are familiar acrivities and may
be comforting (Berkson, 1968). This type of stereotype is to be
distinguished from the locomotor cage stereotypes described by
Meyer-Holzapfel (1968) which are derived from thwarted attempts
to escape. Mason (1967) has developed a comparable theory in
which he uses the term "general motivational state" in reference to
the influence of the degree of arousal on the organization of be-
havior during different periods of development.

In reviewing their own studies of primates, Jensen and Bobbitt
(1968) show how the deprivation of inanimate play objects affects
social behavior.

Inanimate objects constitute another important class of
environmental factors. Our tests of the effects of enrichment
by toys indicate that lack of them and climbing facilities will
seriously handicap the young animal in learning motor skills,
in developing independence from its mother, and in interact-
ing with peers. Further, these effects may be critical in deter-
mining later social dominance. We hypothesized that prior and
concurrent toy experience facUitate peer socialisation.

In an impoverished environment such as a barren cage, there
IS a prolonged period of mother-infant closeness which is essen-
tially a retardaHon in the mutual independence process. Mother-
infant pairs in an enriched environment manipulate themselves
"'"a manipulate the environment more. Jensen

and Bobbitt (1968) propose a conHnuous process of detachment of
infant from mother (and vice versa) as attachments to the larger
environment develop. Apparently deprived infants at 6 months of
age are severely handicapped in social responsiveness when faced
with an ennched peer, with respect to whom they are subordinate.

Stages and Periods in Development

169

These authors emphasize that short-term maternal separation or
isolation from an enriched environment may actually intensify in-
teraction when the infant is returned. This temporary increase in
responsiveness after a brief period of separation (in effect, an in-
stance where absence makes the heart grow fonder) clearly shows
the importance of the duration of deprivation in addition to the
type of deprivation and the age of the organism when deprivation

As in all developmental problems, the role of genetic factors
must be considered first. Henderson (1970) has demonstrated ele-
gantly the interaction between genetic and environmental m u-
ences in the development of mice. Working with several strains, e
found that the results with mice raised in enriched environmen s
were somewhat similar to those obtained by Bennett an ® ^
leagues (1964). The mice were superior to the controls
standard laboratory conditions in performance tests , f •
ious learning tasks and motivational tests of exp ora
Henderson proposes that there is an environmental rep P

uirr^nfplu (and ideally) fi(^

Figure 2. The domesticated phenotype u . ,

loith the type of environment in which different

analogous to a process of pseudospeciatton w ^ ^ ^ conse-

genotypes and phenotypes in the same fitness transitions

quence of different domestic be critical m terms of

between (A) and (B) and (B) and (C) tV

behavioral adaptation.

lyo

THE dog: its domestication and behavior

erating as a consequence of rearing under the relatively im-
poverished environment of the laboratory. He has also demon-
strated that hybrids of the various mouse strains raised under
standard laboratory conditions performed much better than the
pure parent strains in various tests pven to them at maturity (a
typical example of hybrid vigor). When Henderson raised hybrids
in enriched environments, their performance scores were also
higher than those of pure strains that were raised under the same
conditions. Henderson's paper is an important one for everyone
working with laboratory airimals. The extent to which the various
artificial environments in which domesticated animals have been
raised for generations (e.g., home environment for cats and dogs
and "intensive" systems for pigs, poultry, and calves) have influ-
enced brain and behavior remains to be evaluated (see Figure 2).

Socioenvironmental
Influences and
Reproduction

Although imprinting and socialization influence sexual preferences
in later life, other factors relevant to domestication and present
husbandry practices must be considered. Vandenberg (1969), for
example, has shown that the presence of a mature male mouse
accelerates sexual maturation in females compared to females
raised in a unisexed group. There is also strong evidence from wild
animals raised in zoos that being raised together per se (in
heterosexual pairs) may in some way inhibit reproduction. This has
been confirmed experimentally by Hill (1974) who found that pre-
pubertal familiarity in deer mice delays reproduction in nonsibling
pairs. Such a reproductive delay, he proposes, may act to reduce
inbreeding depression and regulate population growth. A third
phenomenon, the social facilitation of reproduction (by visual, au-
ditory, or olfactory cues, depending upon the species) known as
the Darling effect, is pertinent to this review. Animals in social
isolation or in small social groups have a lower reproductive effi-

Stages and Periods in Development

171

ciency than those in larger groups. A male (e.g., a bull) will have a
higher sperm count if given access to females compared to one that
is socially deprived; while females together (e.g., rodents and
dogs) may show a synchronization of estrus."^

These social and developmental influences on reproductive
physiology and sexual behavior warrant further study in domesti-
cated animals and serve to emphasize the close relationship be-
tween neuroendocrine activity and the social environment.

Sensitive and Critical
Periods

Some of the complexities of ontogeny and integrative activity of the
brain and behavior are being opened up for future mvestigation
through these studies that determine the effects of environment^
influences on certain aspects of neurobehavioral development.
Within certain limits, environmental influences can mo^ geneh
cally predetermined patterns of development and moc^ the later
manuring phenotype*. The type and intensity
stimulation may have different effects at different
berg, 1964) and at different ages in different

1968). Similarly, environmental and expenenhal dep"™ may
produce very different effects at various ages ( 3 ““; ^ 9 ^ F°^^"d
fencer 1060! The integration of structuro-functional compo-
n^n'^tHntd associates between different systems mj^al^

be affected (Riesen, 1961) as weU as producing no p
ceneral effects such as influencing the general level of arousa ,
fmotit“and“ response threshold to -“c- or nove^tiniuli
(Denenberg, 1964). Modification of the ® j,

vironmentfl or experiential ZZn-

Ppp^"-'"P “pp" •>"

•Possibly large numbers of animals caged 'n

fertility^ phenomenon suspect in large breedmg kennels

:tty may show a dccrcaw in

172

THE dog: its domestication and behavior

of stimulation or deprivation and also the age and strain of the
species used.

Isolation, deprivation, or selective stimulation and enrichment
may alter the normal ontogenetic sequence of behavior develop-
ment so that certain patterns may disappear earlier with maturity
or persist for an abnormally long time. Meyer-Holzapfel (1968) has
shown that with continued reinforcement, the gape response of
birds may persist and not disappear with maturity at the normal
time. Later developing patterns may not appear at the normal time
because of inadequate or conflicting stimulation or reinforcement,
as shown, for example, by Kovach and Kling (1967) in their studies
of feeding behavior of kittens fed by stomach tube and deprived at
various ages from nursing. To evaluate normal ontogenetic processes
by such techniques may be misleading; what might be shown is the
degree of modifiability or adaptability of the behavioral phenotype
within certain sensitive periods and the dependence (or indepen-
dence) of development and integration upon stimulation, rein-
forcement, and paced increments of experience.

A distinction is now drawn between critical and sensitive
periods because these terms have been used indiscriminately in the
past to describe temporally discrete periods in an organism's life-
time when a number of very different experiences such as handling
(Denenberg, 1964), social experiences (Scott, 1962), and imprint-
ing stimuli (Bateson, 1969) have their greatest and most enduring
influence. This distinction is necessary because some experiences
may be essential for development to continue normally (i.e., ex-
perienhaUy dependent development); whereas other behaviors
and organismic functions may not be dependent upon such influ-
ences for their normal development, yet may be affected at a par-
ticular time early in life by these environmental influences (i.e.,
development is experientially independent but modifiable). A third
category of behaviors and functions may be experientially inde-
pendent and essentially unmodifiable: these are discussed sub-
sequently (see Figure 3).

Critical periods may be regarded as those definable times dur-
ing development when the organism is dependent upon environ-
mental influences for its development to continue normally. Such
experiential dependence at one critical period may represent an
epigenetic crisis (Fox, 1970b; King, 1968), the onset and duration of

MATURATION

CONCEPTION

Structure — ^Function

Figure 3 . Highly simplified schema of various
Hon where Jvironmental influences (E) may or
determining influence, as exemplified by '"^7

lions in domesticated animals. A, environmen a t ^ ..fired" dis-
resistant development, e.g., of barking in dogs J r. recurs
plays. B, environmental

during a critical period, . ofdZelopment may be

such as for song in some birds. C, direc J normal

modified by input during a cLly handling

maturation is not dependent upon this input) e g . ^

stress. D, varying environmental tnpuUunng considerable

may influence maturation in many dirK .
phenodeviance from the norm. (From ox, i9

174

THE dog: its domestication and behavior

the period being species-characteristic and therefore genetically
programmed. There may also be interdependence between critical
periods, where one critical period has an induction effect on a
later period. One example of such delayed effects between experi-
ences during an early critical period and a period later in life, when
behavior is activated and is influenced by experiences during the
critical period earlier in life, is the development of song in certain
birds (Nottebohm, 1970). Another example is the early critical
period for primary socialization which subsequently influences the
development of secondary social relationships (Scott, 1962).

Sensitive periods differ from critical periods in that the or-
ganism is not dependent upon stimulation during such periods for
development to proceed normally. It is rather that at particular
periods in development the organism is especially vulnerable to
certain environmental influences (temperature changes, elec-
troshock, and experimental hormone manipulations) (Morton,
1968) which may have profound effects on subsequent develop-
ment. Experimental manipulations during such sensitive periods
therefore may influence the integration of component parts of a
particular system, e.g., adrenal or gonadal-hypophyseal axis
(Levine and Mullins, 1966), and have long-lasting effects which can
be detected later in life.

In summary, therefore, critical periods may be regarded as
those times in early life when the organism is developmentally
dependent upon certain exogenous stimuli which are normally
present in its sodoenvironmental milieu. Sensitive periods, in con-
trast, are those times when the organism is especially vulnerable to
environmental changes (change in temperature, maternal depriva-
tion), the consequences of which are long-lasting.

The developing organismic system is a link in the intrinsic
complexity of interrelated systems extending in two dimensions;
one in time, between ancestors'and future generations and the
other, in space or place — the organismic system within a complex
of ecological and social systems. The investigator may qualitatively
catalogue, quantify, and dissect development at one of two levels,
at the level of the process itself, or at the level of the interrelation-
ships between processes or systems. Eventually he must reconsti-
tute reality; that is, he must describe the developing organism in
relation to its environment. Then developmental changes in terms

Stages and Periods in Development

m

of geneticaUy programmed, ecologically tuned or adaptive expec-
idenMed”^ human-induced changes via domestication may be

The potential environmental influences (as distinct from gene-
fac selection per se), subsumed under the general phenomenon
designated loosely as domestication, which may greatly alter a
^ven spedes over successive generations are reviewed in this
apter. Early handling stress (or prenatal gentling), imprinting
and sodalization, and environmental enrichment (or deprivation)
during sensitive and critical periods in early life represent the de-
velopmentaUy timed stages when appropriate input may dramati-
cally influence development and later behavior. This is illustrated
schematically in Figure 4 and represents the three key interfaces

Developing and Integrating Nervous System

3

176

THE dog: its domestication and behavior

between the developing organism and its environment where
human domesticating interventions may have their greatest effect.

Emotional reactivity is reflected in autonomic changes in car-
diac activity. This most sensitive indicator has been utilized in a
series of studies to evaluate some of the variables (such as the
effects of early handling stress) discussed in this chapter. A major
consequence of selection for docility in animal domestication may
influence autonomic "tuning" and such possibilities will be inves-
tigated in Chapter 10.

X

Behavior,
Development, and
Psychopathology of
Cardiac Activity in the
Dog

Introduction

Weber and Weber (1845) first demonstrated that ^

vagal nerve causes cardiac inhibiHon. More
(1958) reported that after cutKng the vagal neree, t P P
rate in dogs was almost doubled due to the absence of ^ agal mh.b.

‘‘“"-Belldn (r968) showed that

portant behavioral mechanism m ma y , npctijnn birds,

strated bradycardia Jo^haVre -r dr-ta.led

stud^' M ■t^t- W 3yc”^a-'^fiuring sudde^-rnfi protonr4;fi

immobility in roderits. Vanous reaction, rather

were recorded, and in one spc« • ‘ (,^5-) observed fatal

than tonic immobility, "•‘■'5 p.irasvmpalhetic system,

cardiac arrest apparently mediate^ b) the p^ar. . „ P,„,„„.
in wild Norway rats handled and I

178

THE dog: its domestication and behavior

Bradycardia is also seen in the diving reflex in many diving
birds, in seals, and in man (Scholander et ah, 1942). Gellhom, in
elaborating a useful theory on autonomic tuning of the central
nervous system, observes that

The immersion of the face in cold water leads to a vagal reflex
which increases the resistance of the organism to asphyxia and
is therefore of great biological importance in diving animals.
Studies on this reflex in man with recordings of blood pres-
sure, heart rate, oxygen saturation and skin temperature dis-
close that the immersion evokes an immediate slowing of the
heart rate which persists throughout the test. A rise in blood
pressure and cutaneous vasoconstriction appear after the
bradycardia. This reflex is therefore an example of a (parasym-
pathetic) trophotropic reaction followed by compensatory er-
gotropic (sympathetic) phenomena. If the immersion test is
carried out on an "apprehensive" subject the vagal reflex is
greatly increased, illustrating the enhanced trophotrophic reac-
tivity in a trophotropically tuned subject. Contrariwise, harass-
ing the subject, a procedure known to induce a state of ergo-
tropic tuning, may delay or abolish the immersion reflex, thus
exemplifying reciprocal inhibition of the trophotropic system in
this state.

Other aspects of Gellhorn's theory are discussed subsequently,
relevant to the central issue of the study in this chapter. (Gellhom,
1968).

Sudden fear in man may evoke bradycardia which may be*
followed by fainting (Engel, 1950) or even sudden death (Wolf,
1964).

The excellent review of Fara and Catlett's (1971) study of car-
diac responses and social behavior in guinea pigs provides a basis
for exploring further behavior-related cardiac phenomena in
mammals. Adams et ah (1968) have recorded marked bradycardia
in cats prior to responding to attack by a conspecific.

There may, however, be enormous species and individual dif-
ferences. A normal dog, socialized to people, will show a marked
bradycardia when handled by a person, in contrast to the
tachycardia reported lor the guinea pig by Fara and Catlett (1971).
Does this imply that some species have more or less vagal or sym-

Behavior, Development, & Psychopathology of Cardiac Activity 179

pathetic tone than others? Also, what effects might domestication,
early handling stress (Denenberg, 1967), socialization (Scott and
Fuller, 1965), or the lack of socialization in the dog have on such
cardiac responses.

Heart rate is, to date, the most accessible, if not the most
accurate, measure of changes in autonomic activity — constituting a
quantifiable physiological measure underlying emotional re-
sponses in various social contexts.

Lacey and Lacey (1970) present additional data on the
bradycardia phenomenon in man, as well as some startling, if not
controversial, conclusions. Transient bradycardia is regularly re-
corded during orientation and attention. Those tasks requiring
only simple environmental reception produce significant cardiac
deceleration, while difficult tasks, entaiUng internal cogniHve elab-
oration of a problem-solving sort or requiring exposure to noxious
stimuU, produce massive cardiac acceleration. An activation re-
sponse (getting ready to respond to a conditioned signal), an an-
ticipatory cardiac deceleration, also occurs, which is an indicator
here of high motivation. .

Israel (cited in Lacey and Lacey, 1970) has identified an
ing relationship between this bradycardia of a en on a
"cognitive style" of an individual. So-called sharpeners pay at en-
tion to everything, focus on differences rat er an s‘ ' ‘ '

and habituate slowly (a description that would h' ^

canids). Other human subjects belonged

who tend to make global judgments, habitua ^ ^ .jij;,),

inattentive to many details of the environment^ desmpt on « h ch
partially fits the average domesticated dog). The l^velers sho«^a
lesser attention bradycardia than the sharpener . demon-

of this study is not the analogy with canids, but rather he demon
Station that physiological iabiUty parallels the ° attitude

manifest overtly toward the external envronmen . Centi^^ nert o^us
system activity and behavior <'"''[^‘*'"^j|f,'"f^"greater‘conlrol of
tive style), therefore, appear 'n supposed bv ive.stem

autonomic activity than was hith PP rccocnized

psychophysiologists, although this fac “ 8 Kurtsin's

fho^ugh'^no^ qua^ntified by PaHowan <-^:;:,:u,on-

1968 overview of co^tlCo^^scc^aI ana
ships).

i8o

THE dog: its domestication and behavior

Lacey and Lacey (1970) go on to show that this cardio-
deceleratory response is correlated with Grey Walter's DC shift
in the EEG activity— the contingent negative variation (CNV)
which reflects the organism's readiness to respond. Apparently,
the greater the CNV, the greater the cardiac deceleration.

The question now arises as to which comes first. Research by
Bonvallet et al. and Zanchetticf al. (cited in Lacey and Lacey, 1970)
shows conclusively that cardiovascular afferent input to the CNS
results in a change in CNS activity. For example, elimination of
glossopharyngeal and vagal input (to the bulbar inhibitory area)
leads to poststimulation cortical activation. Thus viscerocortical af-
ferents have an inhibitory effect on the CNS.

Lacey and Lacey (1970) conclude that

The temporary hypertension and tachycardia observable in
acute emotional states and in "aroused" behaviors of all sorts
may not be the direct index of so-caUed "arousal" or "activa-
tion" they are so often considered to be. Instead they may be
a sign of the attempt of the organism instrumentally to con-
strain, to limit, and to terminate the turmoil produced inside the
body by appropriate stimulating circumstances. Moreover, to
quote from an earlier statement of ours,

... if increases in blood pressure and heart rate signal a
physiological attempt to restrain excitatory processes, then it
seems likely that their diminution, absence, or conversion to
blood pressure and heart rate decrease signify an absence of
this restraining process and, therefore, a net increase in excita-
tion; a lowering of threshold, a prolongation of the impact of
stimuli, an increase in spontaneous activity, and the like.

It is this last interpretation, and derivatives from it, that ap-
pealed to us as indicating the most strategic and dramatic ap-
proach to demonstrating the role of cardiovascular activity in
the behavior of intact humans, and to challenging current acti-
vation theory, which has long seemed to us to be a grossly
oversimplified view of the role of autonomic and skeletal activ-
ity in behavior.

(They cite research by Bonvallet et al. showing that distension of
the carotid sinuses could cause a dog or cat electrocorticogram to

Behavior, Development, & Psychopathology of Cardiac Activity i8i

appear as though the animal were sleeping.) Thus^ the possibility
arises that increases in blood pressure and heart rate may be
physiological attempts to restrain excitatory processes rather than
serve as a direct index of arousal or activation. Their logical deduc-
tion of the reverse is that decreased blood pressure and heart rate
signify an absence of this restraining mechanism, a net increase in
excitation, and a prolongation of the impact of stimuli.

Graham and CUfton (1966) have come, independently, to a
similar conclusion vis that an increase in heart rate and an increase
in blood pressure lead to inhibition of cortical activity and would
presumably be associated with reduction in sensitivity to stimula-
tion (as in defensive-protective reaction). Conversely, they
theorized, heart rate decreases should be associated with increase
sensitivity to stimulation (as in orientation or taking in t e envi

ronment).

Lacey and Lacey (1970) also observe that:

We are doing chronic animal studies now 1 *^ ^

strate that thl visceral afferent tal S

operate in the way we think it does, in ® solved

tions we have described. Many problems

and much work done before really effective demonstrations

can be made.

In the meantime, the cardiovascular

psychiatricstudy, notasanonspea cmdexo

tion, but as a h|ghly_speafic ^d appare y
sponse mechanism, integrated at and revealing

affechve and cognitive variaHons P P' ^ j^al ivith

specific personal idiosyncrasies in the way p P j

their external world. (Lacey and Lacey, i 97 - PP-

Sodaliialion of the dog to f"'’" (ho

tachycardia. It was only later that u > ^ ^ dramatic

ining and auscultating the dog was the cause
cardiac phenomenon. nee of a human tvmg

Pavlov (1928) obsen'ed that P , fp^pnnse. so he h.ul lo
could easily inhibit an ongoing conditioned respon

i 82

the dog: its domestication and behavior

isolate his animals for study. Later he found that specific responses
were evoked by certain people, and he postulated the existence of a
social reflex in the dog.

Research on the involvement of the autonomic nervous system
in behavior and in the establishment of conditioned reflexes and
experimental neuroses in dogs was made possible with the advent
of electronic physiological monitoring in the early 1940s. Gantt
(1944), a student of Pavlov, was the first to demonstrate that the
autonomic nervous system was easily influenced by the presence
or absence of human contact, even when no overt behavior was
manifest. One psychotic dog with catatonic symptoms, had a nor-
mal resting heart rate of 200 beats/min, but in the presence of a
human being it would fall to 12 beats/min. Cardiac arrest, some-
times for as long as 8 sec, was recorded in this dog when in the
presence of a person (in spite of which it lived for 14 years in the
laboratory) (Newton and Gantt, 1968).

Subsequent research by Gantt ei al. (1966) demonstrated car-
diac changes in the rabbit, cat, dog, guinea pig, opossum, and
monkey caused by the presence of people. Two consistent observa-
tions were reported in the dog; an increase of 10-80 beats/min when
a person entered the room where the animal was and a decrease of
5-40 beats/min .when petted.

Royer and Gantt (1961) subsequently showed that the cardiac
response to petting in the dog was not an automatic reflex reaction.
Several dogs would only show petting bradycardia with a familiar
handler; a stranger evoked no response or sometimes tachycardia.
Familiarity with the person is therefore of major importance. Scott
and Fuller (1965) noted that there were differences among breeds in
response to a person entering the experimental room, bradycardia
being seen in cocker spaniels and tachycardia especially in basen-
jis. The former breed, especially selected for passivity and crouch-
ing, may be demonstrating a response comparable to fear bradycar-
dia (see discussion later).

In a very original series of experiments. Lynch (1970) demon-
strated that the bradycardia evoked by petting in dogs was not a
simple reflex but was rather an integral component of an emotional
reaction toward the person. He showed that tactile contact has a
potent effect on the cardiovascular system of dogs to the extent that
fear or pain (shock-induced) responses may be blocked, and the
usual tachycardia to such conditioned stimuli was inhibited.

Behavior, Development, & Psychopathology of Cardiac Activity 183

It has long been known that “contact comfort" of a hurt child
(or pet animal) by the parent (or owner) "makes the pain go away."
Lynch (1970) presents physiological evidence in support of this
objectively unlikely phenomenon. The quieting effect of the pres-
ence of handler or familiar keeper in many species of domestic and
wild captive animals is explained by these findings as is perhaps
the potency of reward to a dog in a mere stroke from its master.

Unfortunately, neither Gantt nor Lynch considered the
paradox of bradycardia evoked by contact in dogs that were afraid
of people or unsocialized (feral) and the bradycardia they demon-
strated in socialized dogs that were friendly toward people. A
twitch on a horse or muzzle on a dog may evoke a comparable
calming effect as does the presence of a handler m more tractable
animals. The physiology of restraint and potenHally harmful con-
sequences of excessive autonomic reaction to enforced restraint or
tonic immobiUty induced by pain or fear warrant further study.
Furthermore, genetic influences (selection for domes ca on
addition to such early experience variables (soaalizahon to man)
have not been studied, hence the emphasis m this smdy on wild
(undomesticated) canids and unsodalized (feral) coyote x dog hy

'’"‘^Contact comfort evokes a clear
reflected in bradycardia). A dog will work in or er 8
reward from its Lndler; petting a dog °

pain or react fearfuUy to a condiHoned „

thetic arousal and tachycardia. What role might phen^rncno^
have in development and in establishing soaa
man and animal? , fVm to-

Lynch (1970) reports ,2 and 16 weeks of

sponse to petting develops nod for soanliz,ition

age in the dog, somehme ‘Tfound that the presence of a

(which is from 4-12 weeks ot age). H I rale in a

person (without contact) caused a 40 ,0 , Liddell (1954)

distressed and socially isolated P“PP>', "■'^;|:^'3Vul^n:a mnin’en.
found that if a kid or Iamb was expos .m resist the stress,

in the presence of its mother, it cou Mother succumK'd

sibling hvins stressed in the ab^n dis,urKinci-s into

rapidly, and some showed persislcn

maturity. the lack of tactual

Thus, the question now anscs concerning me

184

THE dog: its domestication and behavior

contact on behavioral development, on the establishment of social
bonds, and on subsequent resistance to stress.

Contact (petting or maternal care) would seem to facilitate the
attachment or socialization process. Deprivation may cause severe
and long-lasting behavioral and physiological changes in man and
other primates (Harlow, 1959; Bowlby, 1953; Spitz, 1950). Tactile
contact appears to be physiologically and psychologically beneficial
(if not essential) for normal development. A dependence upon
parasympathetic stimulation (evoked by the mother or handler)
may also facilitate socialization. In the absence of such other-
directed stimulation, isolated animals (puppies and infant rhesus
and human infants) will effect self-stimulation (tail- or thumb-
sucking, self-clutching, and rocking). The need for such stimula-
tion in immature organisms may be important for the maintenance
of normal physiological homeostasis. Without it, development is
impaired. Also, if deprived of contact, stressful stimuli may not be
adaptively countered by a parasympathetic reaction (as demon-
strated by petting bradycardia [Lynch, 1970]). Stress factors (such
as hospitalization) may then be intensified. Therefore, this is of
considerable clinical significance in veterinary medicine and
human pediatrics. It is also relevant to the husbandry of farm ani-
mals where stresses associated with transportation, weaning, and
maternal separation are major causes of livestock losses and im-
paired growth.

Experiment I.

Development of Heart Rate

The heart rates for four breeds of dog (beagle, wirehaired fox ter-
rier, Shetland sheep dog, and basenji) were recorded at weekly
intervals from 1 to 5 weeks of age. Data from ten subjects of each
breed were then averaged to give an overall view of rate changes
with increasing age. Individual and breed differences were not
analyzed (see later), since the purpose was simply to collect data
representative of a heterogenous, pooled sample.

Behavior, Development, & Psifchopathology of Cardiac Activity 185

Subjects were held in the experimenter's arms and the heart
rate measured using a stopwatch and stethoscope.

From an average rate of 220 beats/min at i week of age, there
was a rapid decKne to 192 beats/min by 3 weeks. After this age, the
resting heart rate gradually increased to an average rate of 208
beats/min at 5 weeks. (See Figure i.)

Suspecting that this developmental pattern was due to a
gradual increase in vagal inhibition, data were collected from sev
eral mongrel and beagle dogs from birth to maturity ?

age), following injection (IM) of atropine sulphate (0.5 mg/ g). is

i86

THE dog; its domestication and behavior

would, by selectively blocking the vagal effect on cardiac activity,
reveal at what age vagal inhibition begins to develop.

Interestingly, and as predicted, there was no significant in-
crease in heart rate until after lo days of age. The greatest release
from inhibition occurred between 2 and 3 weeks of age, which
tends to support the contention that the initial postnatal decrease
in heart rate is related to the development of vagal inhibition.

The more gradual increase in heart rate following atropiniza-
tion from 1 month of age through to maturity suggests that in
addition to the relatively abrupt onset of vagal modulation of car-
diac activity, there is also a more gradual increase in vagal tone.
(See Figure 2a.) This latter phenomenon is to be anticipated in a
species such as the dog, which is characterized in maturity by a
high vagal tone compared to other species such as the rabbit, which
have a greater sympathetic tone. Additional cardiac phenomena in
young caruds are shown in Figures 2b and c.

ONTOGENr or ATROPINE EFFECT ON DOG HEART

Figure 2a. Ontogeny 0/ olropine effect on dog heart.

HEART RATE PER 10 SEC.

Behavior, Development, & Psychopathology of Cardiac Activity 187

mCATUCNT

Figure zb. Various treatment effects on heart rate in normal and
fading 2-day-old puppies: pronounced bradycardia in the latter.
(From Fox, 1966.)

FURTHER DEVELOPMENTAL OBSERVATIONS

Using direct auscultation and biotelemetric monilonn^ heart rale
development was recorded from birth unHI 8 'veeks of age m five
beagles, six coyotes, and three F3 coyote x beag e y n s. n a
subjects, the previously described rate decrease between 2 and 3
weeks of age was confirmed as a developmcnla p cnomcnon in

dependent of handling (petting or human contact).

Petting per se did not evoke a significant bradycardia until
around 5-6 weeks of age; this effect rvas transient b'to m
contrast to the more sustained bradycardia recor e i . - •

Subjects with high resting heart rates "f.
postnatally could be identified later as being the most active and

i88

THE DOG: ITS DOMESTICATION AND BEHAVIOR

I00»»I ,

I5«c.

Figure 2C. Periods of apnoea during various states of sleep are
often accompanied by changes in heart rale.

outgoing canids. Others with lower heart rates tended to be more
passive or timid later in life.

This latter observation compares well with Scott and Fuller's
(1965) developmental study of breed differences in heart rate. They
found that the most outgoing breeds, wirehaired fox terriers and
basenjiS/ had overall higher heart rates than the more passive shel-

Behavior, Development, & Psychopathology of Cardiac Activity 189

ties, beagles, and cocker spaniels. While the fox terriers were con-
sistently different from 2 weeks of age, the other breeds did not
individuate until after 7-8 weeks of age.

Lockwood (personal communication), working in my labora-
tory, concluded that the initial contact bradycardia of short duration
in very young canids may be analogous to an orienting response
(see also Lacey and Lacey, 1970). He also recorded a transi-
ent bradycardia when contact with a neonatal canid was broken;
this “off response" may be an important autonomic indicator re-
lated to behavioral regulation of homeostasis, i.e., contact with
mother for thermoregulation and nurturance must be maintained.

Confirming the findings of Scott and Fuller (1965), it was
found that heart rate and sinus arrhythmia are highly interdepen-
dent in older canids. A slow heart rate (i.e., high vagal tone) is
arrhythmic, and conversely, a high rate is regular without any
arrhythmic inclusions. Scott and Fuller (3965) demonstrated breed
(genetic) differences in these interdependent physiological indi-
cators of autonomic activity and emotionality. It will be demon-
strated shortly how early experience (handling stress) may also
influence cardiac activity in later life.

Experiment 11. Can Early
Experience Influence Cardiac
Activity?

Research on early handling stress in rodents (Denenberg, 1967;
Levine and Mullins, 1966) has shown that the adrenal-pituitaiy
axis can be affected so that reactions to physical or emotional stress
are more graded and adaptive than in control animals. The latter,
unstressed in infancy, tend to overreact to stress. Early handling in
rodents also enhances learning ability under stressful conditions,
since emotional arousal is loss than in controls and so does not
interfere with performance. Resistance to cold exposure, starva-
tion, implanted neoplasms, and wal leukemia has also been
shown to be greater in rodents subjected to optimal handling stre^^s
in early life. Even more remarkable is that this phenotvpic

190

THE dog: its domestication and behavior

modification is transmitted nongenetically over several generations
(Denenbetg and Rosenberg, 1967) and may be instigated prenatally
by handling the pregnant mother (Joffe, 1969).

The findings imply that environmental influences (i.e., stress)
during a critical period in early life have profound and enduring
consequences and may well influence sympathetic-parasympa-
thetic tone as well as determine emotional reactivity or tempera-
ment in later life.

Blizard (1971) has discussed the significance and possible clini-
cal implications of individual differences in autononaic responsive-
ness (cardiac activity) in laboratory rats. He found that rats given
handling stress from 1-7 days of age exhibited dimiiushed cardiac
response to handling and to novel stimuli in adulthood compared
to nonstressed controls exposed to these stressors for the first time
as adults. Environmental influences early in life, therefore, have a
clear effect on emotionality and, because of the reported influence
on heart rate, may enhance trophotropic or parasympathetic tone
in laboratory rats.

Theoretically, at least, the autonomic-neuroendocrine interre-
lationships and temperament per se may be similarly modified in
the dog. The potentials for such phenotyq>ic engineering may also
have considerable practical value in raising dogs for stress resis-
tance, an important consideration for military and peace-force use.

Given, then, that vagal tone is not already established at birth
in the dog, it might be possible to modify sympathetic or parasym-
pathetic tone by enhancing the former via handling in early life, as
demonstrated in the aforementioned studies with laboratory 10=
dents.

Therefore, it was decided to evaluate the physiological and
behavioral effects of early handling in domesticated dogs. Four
litters of random-bred pups were used, half of each litter being
used as nonhandled controls (eight in all) and the remaining eight
pups subjected to the following regjme from birth until 5 weeks of
age.

MATERIALS AND METHODS

This experiment was designed to determine the effects of differen-
tial rearing on several aspects of behavior and development of the

Behavior, Development, & Psychopathology of Cardiac Activity 191

dog. Sixteen dogs were studied (eight control, eight handled). The
handling procedure and behavioral tests are described in Appen-
dix I.

RESULTS

No significant difference in body weight gain was observed in the
groups, nor were significant differences in organ weights or in total
brain weight observed. When subjects were tested at 3 and 4 weeks
of age by a variety of neurological responses (Fox, 1964), no signifi-
cant differences were observed. Some handled pups, however,
showed slightly superior coordination while standing and walking
at 4 weeks. Histological examination of motor, occipital, and fron-
tal cortex and vestibular neurons revealed inconsistent differences
in neuronal size, but no differences in cell density were observed.
Vestibular neurons in four handled subjects were significantly
larger and contained more chromatin than those in the control or
isolated pups; the total population of neurons observed in senal
sections contained fewer small-sized neurons. The Meynert cells of
the fifth layer in the occipital cortex and those of the fifth layer in
the frontal and auditory cortices (large pyramidals) appeared larger
in three handled pups compared with Uttermate controls, and they
contained more Nissl substance. Heart rates were recorded in the
control and handled groups at weekly intervals
at 5 weeks of age with surface electrodes, at which tune su^ects
were not handled (see Figure 3). Tlie marked

the handled and control groups were apparent from the second
week onward. Normally there is a decrease in heart

rdLtssIdTarhTw^^^^^^^^^

found to be 63% ± 7 m the contro . Hlinc^ but there was a

total amino content of the adrenals due 1 ende con-

significant increase in epinephnne. n in g

tent of the kidney and adrenals controls, which

found in the handled pups compared with their

192

the dog: its domestication and behavior

SLEEPING EEG OF HANDLED ANO CONTROL 5 'WEEK OLD PUPS

HANDLED

CONTROL

ERG— 4444444444444^

so pvl I

l-SEC.

•4>-44'4444J’-44^44'J’

Figure 3. Sleeping EEG of handled and control g-week-old pups.
Note greater maturity of EEG of handled pup (higher amplitude)
and also faster resting heart rate.

may indicate an increase or change in lipid metabolism as a result
of early handling.

BEHAVIORAL OBSERVATIONS

Extreme differences were not found among individuals in the same
group, and this surprising uniformity facilitated comparisons be-
tween the two differentially reared groups. These data have been
summarized in Table 1. Generally, the handled pups were hyperac-
tive, highly exploratory, very sociable toward humans, and domi-
nant in social situations (e.g., play) with their peers. The handled
subjects showed the greatest distress vocalization immediately after
the handler had entered the testing arena and removed the cloth
and toy. In contrast, the control subjects were little distressed by
this interference but showed great emotional arousal when first put
into the arena. In the barrier test situation, the handled pups per-
formed best in that they requited fewer trials to negotiate the dc-

Table I.

Behavior, Development, & Psychopathology of Cardiac Activity 193

194

THE dog; its domestication and behavior

tour, whereas the control pups reacted more slowly and showed
distress vocalization in this situation (Table I). In the control group,
therefore, emotional arousal in this situation prejudiced problem-
solving ability.

Recordings of EEG of all the handled pups showed a greater
amplitude during light sleep than those of the control subjects
(Figure 4). As amplitude increases with age, it may be presumed to
be an indication of greater maturity in the handled subjects. No
difference in amplitude or fast frequency components were ob-
served between the handled and the control pups in the alert state.
A similar observation has been made on EEG maturation in kittens
following early handling (Meier, 1961).

From these findings a pilot project was set up in cooperation
with the Biosensor Research Project, U.S. Army Veterinary Corps.
Several litters of German shepherd pups were given the early han-
dling stress from birth to 4 weeks of age, each subject receiving
only 1 min on a 45 rpm teeter-totter that gave both lateral and
angular vestibular stimulation and i min cold exposure at 37°F.

Rgun.- .(. Hart ralfS of conitol and handled dags. Stale sustained
higher hcnrt rales in handled pups.

Behavior, Development, & Psychopathology of Cardiac Activity 195

Prelimmary results so far are extremely promising. As adults, such
dogs have reportedly superior stamina under conditions of heat
stress in the tropics. Data on disease resistance and longevity are
not yet available. But a word of caution should be added. As
Ginsburg (1968) has shown, working with different inbred strains
of mice, the same handling stress may vaiy from one strain to
another — having no demonstrable effect in some, enhancing oth-
ers' later stress resistance, or lowering stress resistance in others.
In a more heterogenous dog population, therefore, early handling
may produce comparable results; thus, the quality and quantity of
stress must be carefully regulated on an individual basis. Ideally, a
suboptimal handling regime should be instigated (e.g., i min
teeter-totter stimulation followed by 1 min gentle stroking, then 1
min cold exposure followed again by 1 min gentle stroking).

Under natural conditions, a wild canid mother may leave the
cubs for extended periods, and they would be exposed normally to
more stresses in early life than the relatively overswaddled home-
raised litter.

Several breeders over the past 2 years have set up a simple
handling stress regime using half of each litter, the other half being
controls. Based upon subjective reports, all are enthusiastic and
believe that this method of phenotype enhancement Is a valid phe-
nomenon (which controlled laboratory research has already dem-
onstrated!) and a valuable addition to the usual rearing practices.

In fact, there is less originality in the application of this phe-
nomenon in improving the temperament of domesticated animals.

As Denenberg and Whimbey (1963) have suggested, prenatal han-
dling or gentling of the pregnant mother and early postnatal han-
dling of the offspring were probably practiced unwittingly as a part
of animal husbandry early on in the domestication of both farm
and pet animals.

The marked difference in the age-related changes in heart rate
in the four breeds of dog described earlier and the control subjects
in this latter study warrant scrutiny and explanation. The controls
received regular daily contact with people and were well sodalizcd
and more used to being handled than the pups of four different
breeds that were handled only at weekly intervals for recording
growth and heart rales. The increase in heart rate in these dogs
after 3 weeks of age (at the beginning of the critical period of

196

THE dog: its domestication and behavior

10 sec

Petting Bradycardia ^ Dog

Figure 5. Petting bradycardia in a female dog.

socialization) may be a reaction to being restrained and handled.
Such tachycardia, indicative of flight or emotional arousal (fear),
has been recorded in several dogs of various ages. In socialized
canids, however, a slowing of the heart rate occurs (see Figure 5).
This contact bradycardia is discussed subsequently. It should be
emphasized that the tachycardia that developed in the handled
pups after 2 weeks of age was not correlated with fear or other
emotional disturbance. On the contrary, these "super-socialized"
puppies never resisted restraint or handling. Most significantly,
they also manifested tachycardia even when resting unrestrained,
with EKG being recorded by surface electrodes rather than by
stethoscope (see Figure 4).

Clearly, such variables as socialization, habituation to han-
dling, restraint and the method of measuring heart rate (either with
surface electrodes to a polygraph or.via stethoscope) are significant.
Subsequently, therefore, heart rates were recorded with a bio-
telemetry transmitter directly through a receiver onto a physio-
graph.

Experiment III.

Individual Differences in
Heart Rate and Temperament

Tile next question was to follow up the marked effects of early
handling on cardiac activity and behavior (i.e., phenotype
modification) with a study of individual differences within litters of
canids given no handling stress. Individual differences in heart

Behavior, Development, & Psychopathologi/ of Cardiac Activity 197

rate, if present, may correlate with other measures of emotionality
and behavior and be an indicator of autonomic tuning or sympa-
thetic tone independent of environmental (stress) influences; such
individual differences, which can be created by handling stress,
may also be genetically predetermined as part of a young animal's
basic psychophysiological constitution or temperament.

If such individual differences could be demonstrated, how en-
during might they be and how, under natural conditions, could
they influence disease resistance, fertility, and other eventualities
demanding adaptation to some form of stress?

Also, if there is consistency in such individual differences in
heart rate and a high correlation of this measure of autonomic
activity with other measures (of behavior, emotional reactivity,
and learning abiUty), we may have an extremely sensiHve index or
test protocol for puppy evaluation. Such early evaluation has ^eat
applicability for screening litters of puppies and selecting in lyi u
als at an early age for special rearing and training programs (e.g.,
military dogs, guides for the blind, etc.).

This possibiUty has been explored from a somewhat differenl
perspective in man. Thomas el al. (197°) devise a °

tests to evaluate the emoHonai/autonomic
neonates. From these measures, they were ® ^ ° ament would

predictions as to what the later personality or . ^dv the

he-, these predicHons were verified in th.s

subjects now in their late teens. From a prac ” ^ . ^^^jd

relative to dog and man, such early ^ jp order

«n.bl. one to In.lijele ,h„

emotional disturbance following change m rou
of immediate reward or comfort). . , caotive adult

An initial pilot study --

wolves at the Naval Arctic Rere ^ b.otelcmeters

Alaska. Working in collaboration I ^ yearling wolves, the

were implanted bers of adaptive pack being

highest- and lowest-ranking m j ^g hr to monitor

selected. Recordings were '^ben

possible circadian vanables m ac > highest-ranking wolf

and overall motor acliv'ily was grea

THE dog: its domestication and behavior

198

It was tentatively proposed (Fox, Folk and Folk, 1970) that such
marked physiological differences might be predetermining factors
for temperament and social rank.

With these promising findings, the following experiment was
designed to explore further the possible significance of individual
differences in heart rate correlated with other behavioral measures
in unsodalized 6 to 8-week-old wolf cubs. This species was chosen
in order to rule out any potential anomaly in vagal tone which, in a
domesticated species, together with socialization effects, might
have a significant influence on cardiac activity.

In the present investigation, a temperament or reactivity pro-
file was made for each cub in two captive litters (as detailed by Fox,
1 ^ 97 ^)/ and the social rank of each cub was ascertained. Cubs of
high, low, and intermediate scores were then selected for
biotelemetric monitoring of heart rate under various test condi-
tions; some of these subjects were then used in a stress study, the
response being determined by changes in plasma cortisol levels
over time.

MATERIALS AND METHODS

Twelve woiS cubs from two litters were studied between 6 and 8
weeks of age. A series of tests, described in detail in a recent study
of individual differences in behavior of wolf cubs (Fox, 1972), were
run (see Appendix II).

RESULTS

The scores for group and individual prey-killing, dominance (aver-
aged scores from repeated tests), and the reactions to the novel
stimulus are detailed in Table II. The close correlations between
social rank, prey-killing, and exploratory behavior, as demon-
strated earlier (Fox, 1972) are clearly evident in these two litters. In
this earlier study, the question of social facilitation versus prior
experience was posed since the group prey-killing test was con-
ducted after the individual tests. The data in Table 11 show that
social fadlitation may indeed be present in the group test and

Behavior, Development, & Psychopathology of Cardiac Activity

Table H.

SCORES FOR GROXJP AND INDIVIDUAL PREY-KILLING, DOMINANCE, AND
REACTIONS TO NOVEL STIMULUS

Group Individual Exploratory

prey-killing prey-killing Dominance behavior

Litter I

ScBl 6^ 5

BI 5

ScBr 6 4

RcBl 9^ 4

UcBr 9 4

CCBr d 4

CCBr 9 4

KcBr d* 4

5

5

4

4

5

3

4
0

5

4

4

3

3

3

3

2

5

4

5
5
5

4

5
0

Lii/CT- II

ScBl 2 3

RcBl 62 1

BlUc 9“ 1 0

CCBl d" 0 0

•Selected for EKG and corticosteroid stress studies.

would account for the high scores of

tested alone later, despite prior expe^nce, low sco^Dd
ferences in the magnitude of scores behveen ^ ,

marked. In the earlier study O^ox, ’97=)'
produced a similar outgoing litter the pr • ^

lowest scoring litter in this ' '

dam to Utter II (both being sired by pmeedur.-s

Changes in heart the vanous h^ndh^^^^^

aregraphically represented m F^ 6 ^nuct and

finding was the bradycardia assoa^ r -

Heart rote /mm

200

THE dog: its domestication and behavior

orma; j. b 4. canlacis the aib; 5. obfcrvcr Imb off; 6. cub is picked
up and held; 7. cub is ieft in arena and baseline rale recorded 8
ol^rrer makes eye contact; 9. baseline rate recorded; 10. response
to suddai noise.

passive submission (or freezing). A second consistency was the
lugher baseline heart rates in the dominant cubs over the subordi-
nates, a consistent:)' which adds to the obsers'ations of Fox et al.
(1970) who reported a higher heart rate over a 24-hr circadian
periodicity study of a dominant yearling in contrast to its subordi-
nate littermate. Bradycardia also occurred when the cubs were
picked up. A less marked bradycardia was recorded when the ex-
perimenter rem.ained passive in one comer of the arena and also
when eye contact alone was made around the .screen without the

Behavior, Development, & Psychopathology of Cardiac Activity

201

landn

experimenter being in the arena. In toth cubs. Tachycar-

eye contact was most marked in the o recorded in RcBI ^

dia, instead of the anticipated g and 7 ), ihis reaction

ofLitterlandCCBl d of Utter H (see Ft^res^=^^7;

being correlated with defensive-aggro® those cubs ®o*’)ocle
Changes in /xg% cortisol U tho p a .f^gnt trends (see a c

to the stress experiment revealed som [,jrh arc quite capa c

HI). Generally, the highest-ranking^' delay the re-

of adrenal response (according to bccon. lo contrast,

sponse for a time after the ^

of intermediate social rank resp<^ f further secretion

of cortisol secretion and are capab eo in te

treatment. Subordinate animals am n jtn.’S'^. oor do tht>
increased levels of plasma cortiso ^ ACTM

a capacity for adrenal secretory resp

202

THE dog: its domestication AND BEHAVIOR

Table III.

SUMMARY OF WOLF ADRENAL RESPONSE TO STRESS

Cortisol in plasma (fig%)

B.W.

Animal

Condition

(Ibloz)

5 min

10 min

30 min

60 min

Utter I

a ScBl d

control*

1516

22

70

61

36

/3B1 9

control

1513

39

69

190

65

e RcBl 9

control

15112

95

100

88

50

(i> KcBr d

control

16/8

38

25

24

26

Utter II

a ScBl 6

control

14114

15

38

25

48

€ BlUc 9

control

1210

123

26

24

18

u CCBl d

control

13/13

22

20

20

36

Utter I

a ScBl d

ACTH"

1516

92

100

184

47

to RcBr d

ACTH

16/8

23

24

26

41

Lillcr 1/

o ScBl d

ACTH

14/14

35

50

20

30

€ BIUc 9

ACTH

12/0

60

36

25

34

•Control, serial blood samples taken from time of confinement with no further
treatment.

‘ACnf. troated with ao lU Acthar (IM) at time of conrinement.

It should be emphasized that the cubs of Utter II were more
timid than most of those of Utter I (see Table II); in the responses of
the dominant cubs of the two litters, for example, the dominant
cub of Utter II resembled more a lower-ranking cub of Litter I.

T^ese findings are discussed subsequently and correlated with
individual differences in both cardiac responses and behavior ns
demonstrated in the various tests described earlier.

Behavior, Development, & Psychopathology of Cardiac Activity 203

DISCUSSION

The behavioral test scores for prey-killing ability and exploratory
behavior show a close correlation with rank scores for social domi-
nance, especially in the more outgoing Litter 1. These findings
support the earlier study of individual differences in wolf litters of
Fox (1972). In this earlier study, the question of social facilitation of
prey-killing arose and was not completely answered. The present
study was in part designed to answer this question, by giving the
test v\dth live prey to the whole litter prior to testing each indi-
vidual. Several cubs had lower scores when tested alone with live
prey than they had when tested with conspedfics.

The long-term significance of these tests should also be em-
phasized; cubs tested at 8 weeks could be identified ^ the basis of
their scores when retested t year later (Fox, 1972^ This serves to
highUght the notion that the temperament of each individual es-
sentially is determined innately. Or, at least f e
phenotype is formed by 8 weeks, so that by this age ear y exp
ence has established a stable state (i.e., ' •

teractions are stabilized by 8 weeks). This asic e P '

which may be a constellation of autonomic
pathetic response threshold, and emotional 5"^'

influences the way in which ivjih

to prey, and to novel objects. Such expenence , critical

a particular temperament constellation during hchivior at

wLks of life, determine subsequent =°7'>'.”"Vwhh^^hrpack
least up to the first year of life. Certain wUhm p^ick

may confound such predictions where a -aj

become a subordinate outsider with ma un y . hichcr rcst-
A consistent finding in the heart h^h-

ing heart rate, indicative characteristic of the

ranking wolves, while '°'r''j2nKS support those of Fox el at
most subordinate cubs. These fin g .PP ,j ,j, (in a arcadian
(1970) who found a higher heart rate nd ac .
analysis of cardiac activity) in dominant tn

ling wolves. . nne parts of the test se-

The marked bradycardia ma fn-ezine or passive im-

quence with a handler is associated with fn-czang P

204

THE dog: its domestication and behavior

mobility and has been recorded in many vertebrates (Belkin, 1968),
as well as in unsodalized F, coyote x beagle hybrids during han-
dling (see later). This bradycardia is to be distinguished from pet-
ting bradycardia in sodalized dogs (Lynch, 1967) and other canids
(see later). The contact bradycardia is of much longer duration than
the orienting and attention bradycardia described in man (Lacey
and Lacey, 1970) and is to be differentiated from the bradycardia
which occurs after stress-induced tachycardia as immediately fol-
lowing a fight, for example (see later).

Candland et al. (1970) have studied changes in heart rates in
squirrel monkeys in various sodal contexts, notably during aggres-
sion and establishment of dominant-subordinate relationships.
They found that heart rate was related to rank on the status hierar-
chy by a curvilinear function; middle-ranking animals showed the
lowest heart rates during test sessions but not during baseline
measures in the home cage. In chickens (Candland, et ah, 1969),
they found a similar U-shaped function; alpha and omega birds
showed the highest heart rates during intraspecific test sessions,
and middle-ranking ones had the lowest rates. These authors ques-
tion why animals of both high and low rank should show higher
heart rates than middle-ranking animals. It is not inconceivable
that those of low and high rank have a greater sympathetic arousal
associated respectively with threat or readiness to attack and flight
or defense. Those of middle rank may have a greater parasym-
pathetic tone related to passive submission or passive avoidance of
conflict, so that escape-assodaled tachycardia is not manifest as in
subordinates, (i.e. Tachycardia may be related to fight and flight in
high- and low-ranking individuals.) Marked tachycardia was re-
corded in the squirrel monkeys during handling as distinct from
the bradycardia of the wolf cubs.

Candland et al. (1970) found that when an individual experi-
enced a change in social rank, there would be a predictable change
in heart rate which correlated with its new rank position. These
findings would be interesting to follow up and compare with
wolves which might also show comparable physiological as well as
behavioral changes as a consequence of a change in social relation-
ships.

Murphree et al. (1967, 1969) reported lower resting heart rates
in a timid strain of pointers, while a more stable strain had a higher

Behavior, Development, & Psychopathology of Cardiac Activity 205

resting heart rate. These findings compare well with the test scores
of high- and low-ranking cubs which are analogous to the timid
and stable temperament types that Murphree et al. studied. These
authors were only able to demonstrate bradycardia during petting
in the stable strain, a phenomenon associated with the reactions of
socialized dogs to familiar handlers (Lynch, 1970; Lynch and Gantt,
1968). Bradycardia has also been reported in vertebrates in re-
sponse to threat (Belkin, 1968). Again we have a paradoxical phe-
nomenon; bradycardia evoked by both threat and petting but
associated respectively with passive immobility or freezing and pas-
sive submission, both behaviors possibly involving parasympathe-
tic relaxation and cardiac deceleration. The bradycardia associated
with threat or stress may reflect the animal's readiness to respond;
a vigilance state which has been correlated with changes in elec-
troencephalographic activity (Lacey and Lacey, 1970)-

Gellhom (1968) has proposed that differences m temperament
and behavioral reactions may be attributed to differences in sympa-
thetic (trophotropic) and parasympathehc J?,

tone. Schneirla's (1965) promising but as yet bt e es

of biphasic sympathetic-parasympathebc processes jj

togeny and organization of behavior is also re pvolain why

erties. Thomas et al (1970). for example, ha temperament

dons between the behavior of =>dotescen.s and the r ^
consteUations that were determined by a battery of S

first few months of life. differences in nervous typolog)-

The Pavlovian concept of ddferen^

(synonymous with ‘emperamen ) (^r dynamism), and

Typologies differ in terms of bala, and^ excitation. Thus, a

relative strengths noloEV or temperament ivould be

strong balanced but dynamic tyP ^ , pss (neither overreactive
characterized by adaptive respon balance between sympa-

nor hyporeactive) where the horneo regulated (Kurtsm,

thetic md parasympathetic systems is «eli gu

2.o6

THE DQG* rrs DOMESTICATION AND BEHAVIOR

1968). Such an animal would show appropriate behavioral reac-
tions to tests requiring passive inhibition, approach, or with-
drawal.

The data on plasma cortisol secretion under stress and capacity
to respond after ACTH treatment as well as the heart rate data
accord well with the above concepts. It may be tentatively con-
cluded at this level of analysis that dominant wolf cubs have a high
sympathetic tone but do not overreact to stress (i.e., parasym-
pathetic homeostatic regulation or balance is evident). Low-
ranking cubs have a low sympathetic tone and show a hyporeactiv-
ity to stress. Those of intermediate rank have a higher sympathetic
tone but tend to overreact to stress.

Follow-up studies over a 3-year period confirmed that these
individual differences in temperament and rank order, established
very early in life, were enduring provided the relationships with
the Utter were not disrupted (as by the death of one key individual
or the introduction of a stranger).

These findings indicate that heart rate is a reUable index of
temperament, but since both are determined geneticaUy and expe-
rientiaUy, changes in social relationships may affect heart rate. This
was not determined in canids, but the work of Candland et al,
(1970) with squirrel monkeys shows that a change in rank is corre-
lated -with a change in heart rate. They observed that

Of special importance is the fact that when animals changed
position on the (rank) order their heart rates changed accord-
ingly, maintaining the curviUnear relationship. The impUcation
of this result is that when dominance orders change, the heart
rate of the animals changing position also changes. It suggests
that rank determines the heart rate, rather than the reverse.

This latter conclusion is unwarranted however without a de-
velopmental study; their subjects were apparently unrelated and
from diverse sources.

Candland et at. (1970) found that there is an inverted U- or
J-shaped curviUnear function with the lowest resting heart rates in
middle-ranking animals, in contrast to the wolf where rank corre-
lates more directly with heart rate.

These authors ask, since they have also demonstrated a very
similar correlation between rank and heart rate in chickens

Behavior, Development, & Psychopathologi/ of Cardiac Activity 207

Why would animals both high and low on the order show the
higher heart rate? Or, conversely, why should midranking
animals show the lowest heart rate? One possibility is that the
noradrenalin/adrenalin ratio differs substantially in alpha and
omega animals and has the result of generating different forms
of behavior (flight or fight) while increasing the heart rate in
both cases.

Since neither of these species are predators, the observations
of Funkenstein (1955) might account for the differences in cardiac
activity between wolves, squirrel monkeys, and chickens, tie
states that

Experiments suggest that anger and fear may acHvate different
areas in the hypothalamus, leading to production of nor-
adrenalin in the first case and adrenalin in the secon . n
more e^eriments are made, these possibilities mus remai
suppositions.

Some of the most intriguing work in this field
ported by von Euler. He compared adrenal secretions found in
a number of different animals. The researc ™ adrenal

supplied by a friend who flew to Afnca o Ruesch

medullae of wild animals. Interpefing his

ou. ,h., 7"^ “‘h K • rL

atively high amount of nor-adrenali , flieht ad-

the rabbiCwhich depend for survival Xals

renalin predominated. also have a high

that live very soaal hves (e.g., the
ratio of adrenalin to nor-adrenalin.

It may be, therefore, that exhibit

noradrenalin (associated with aggressi j those with a

tachycardia in socially o. assertive

greater sympathetic . ^5^ ^^ith a lesser sympathetic

and assume a higher rank should also be emphasized

tone (and lower resting heart rat ^ ^ social reaction

that passive inhibition rather than be correlated w«th

to threat in low-ranking was evaluated in a

a lower resting heart rate. This speculation
subsequent study (see later).

2o8

THE dog; its domestication and behavior

Contrary to the notion that high vagotonia is an attribute of
athletes and of dogs with superior physical stamina, what adaptive
value would high sympathetic tone have for the wolf? The higher
sympathetic tone experimentally induced in domesticated dogs
(described in Experiment II) showed that such animals were emo-
tionally more stable, more exploratory, and assertive. In the social
context of the pack, such an individual would be more likely to
survive under conditions of food shortage in a severe winter, and
such wolves are usually the only ones to breed in a pack. How
individual variance or heterogeneity is maintained in litters of off-
spring to counter this unidirectional selection is an important ques-
tion and deserves further study.

As Linn (1974) emphasizes, "the relative effect of ACTH and
plasma corticosteroids on the psyche of the animal may well be
determined by the state of the ergotropic-trophotropic balance of
the hypothalamus (GeUhom, 1968). Gellhom (op. cit.) has
theorized that ergotropic-trophotropic balance is closely interre-
lated with secretion of adrenal corticosteroids. Increased cortico-
steroids reduce sympathetic hypothalamic reactivity, cause a con-
current rise in parasympathetic hypothalamus reactivity and
homeostatic balance is shifted to the trophotropic side."

Field observations show that the lowest-ranking wolves are
the least likely to survive a severe winter (again reflecting the find-
ings in nonhandled rodents being less resistant to terminal starva-
tion or cold exposure than those handled subjects that have an
induced high sympathetic tone). The data on plasma cortisol levels
and response to ACTH imply that such canids are essentially inca-
pable of mobilizing an adapdve stress-response. This contrasts the
overreaction to stress in nonhandled rodents; it may also account
for the low resting heart rates in low-ranking carnivores, and such
prey spedes of comparable sodal rank (as studied by Candland et
al, 1970) may have abnormally high resting heart rates.

Behavior, Development, & Psychopathology of Cardiac Activity 209

Experiment IV. Heart Rate and
Plasma Cortisol as Predictors

of Temperament

The following is an abstract of an extensive study by my colleague
and student. Dr. J. Linn, representing an attempt to employ the
testing procedures used in controlled laboratory conditions to a
more uncontrollable kennel-rearing facility operated by army per
sonnel. (For details of test procedures see Fox, 1968, 1971; Linn,
1974; and Appendix III.)

The plasma cortisol level and heart rate of one hundred and
rune twelve week old German Shepherd Dog puppies was
monitored prior to, during and after undergoing a
cally stressful experience. The physiological parame e
then examined for correlation with temperamen o g

as they matured.

Because of large standard deviations
significant correlations between either of t e pa
sumd and temperament. Plasma cortisol
trend of prolonged elevation in * g j

pies with the poorer temperament. (See igu

There was a consistency in barline and fif-

nations. The mean plasma ’^hape with the highest

teen minute post-stress showed a P of the J.

rated pups and the lowest rated pup highest cortisol

Prior to testing the top rated puppies had ™ j
levels and fifteen minutes after testing the lowest P
pies had the highest cortisol leve s^ ^

The mean heart rate of the group P mean heart rate

temperaments did not increase as mu stressful situation,

of the passing group when Pf ’"'“ir^s was higher than
Their mean heart rate^o mm P showed trends to

in the passing group. Thus, P ,gi,o!ogical stress m fail-
prolonged physiological P ^ standard deviation on

ing puppies. However, the rang

210

THE dog: its domestication and behavior

I Standard Error of AAean

(23 Baseline, Cortisol
r~~l 5 ^Ainute, Cortisol
M 15 Minute, Cortisol

TEMPERAMENT SCORES

Figure 8. (From Linn, 1974.)

cither side of the mean for passing/failing groups overlap too
much to allow successful prediction of future temperament.
(See Figure 9.)

A critical reevaluation of this potentially very promising
screening procedure for early puppy selection reveals one impor-
tant methodological omission. Resting heart rates were never

Behavior, Development, & Psychopathology of Cardiac Activity

211

monitored biotelemetrically from jnjHal jueular vein

Base heart rates were recorded f "Jn^'keMsee Fig-

sample for plasma cortisol detenrunaho
ure 9).

212

THE dog: its domestication and behavior

This prior stress served to segregate the pups on the basis of
an inverted U function. The two groups of pups scoring best on
temperament ratings (types 2 and 3) being intermediate between
the reject pups (types o and i) which showed relative bradycardia or
tachycardia poststress. The greatest increase in heart rate occurred
in the highest-ranking pups when placed in confinement (Box)
prior to release into the novel environment (Arena). All four
groups of pups showed a similar increase in heart rate in the arena.

The most significant finding is the heart rate recordings taken
2 min after exposure to the novel environment (i.e., recovery).
Return to a resting level was graded inversely in relation to the
temperament rating, the lowest rating pups having the highest
heart rates, reflecting a sustained tachycardia following stress.

As in previous studies involving bradycardia in animals, insuf-
ficient distinction of behavioral signs was made between the
bradycardia of fear and of being petted or restrained by the han-
dler. Both friendly and fearful dogs may show passivity (passive
submission); monitoring of muscle tone (EMG) may help dif-
ferentiate these two states of passivity in future studies.

An additional factor which naturally reduced the variance be-
tween categories of test subjects was in the initial selection proce-
dure; extremely timid pups were discarded initially and were never
used in the study. Had they been included, a more natural,
heterogenous sample could have made a significant difference in
the final analysis of the data.

Experiment V. Bradycardia in
Fear and Friendship

Bradycardia was evoked consistently in unsocialized (feral) canids
(see Figure 10). A 20-30% decrease in heart rate from "resting-
alert" was usually recorded. The normal baseline of resting-alert
was taken, since general locomotor activity causes a considerable
increase in heart rate.

Resting-alert heart rates were recorded biotelemetrically with
the subjects alone in the testing cage and monitored via closed

HEART RATE 250/mi n

Behavior, Development, & Psychopathology of Cardiac Activity 213

Figure 10. Heart rate decrease with contact in unsocialized coy
X dog hybrids.

„ aUn recorded when a handler
circuit television. Heart rates were subiect, remaining

entered the cage and sat 6 ft the handler sKxid

passive and avoiding eye contact. ^ , physical contact (pel-

up, slowly approached off and left the cage F.ve

ting) for 1 min, and then slowly b

repetitions for 3 consecutive <i3ys ^ ceneration coyofe >

In the four unsodalized canids s subjects dunng

dog hybrids), bradycardia devel<^ represent a conditioned

repeated tests preceding Ficure n )

autonomic response or expectation.

214

THE dog: its domestication and behavior

SC02 Sitting

Approach

'Freezes'

Contact

* 5 sec *

Figure ii. Precontact bradycardia.

Tachycardia was frequently recorded when the handler ini-
tially entered the cage and later when he stood up preparatory to
approaching the animal. In the latter context, no subject attempted
to escape although this cardiac response may reflect a readiness to
escape — the flight response was blocked by the confines of the
cage. In one subject, tachycardia was recorded as the handler
withdrew atter making contact; this was correlated with attack
intention movements (presumably evoked by the flight of the han-

^ CD! Sitting

Abroach Contact —

ww

Contact continuad

Figure 12 .

Behcmior, Development, & Psychopathology of Cardiac Activity 215

Figure 13.

dler). Over consecutive tests in the other ^^I'Z^p'fj^gbitu-

approachgraduallydisappeared — aphysiolopca -ve contact

adon (see Figures^ra and 13). When the handler made eye
with the subject, a transient bradycardia was X

(see Figure 14).

I *

' 70 fC

Figure 14.

2i6

THE dog: its domestication and behavior

hrodyccrdta.

These cardiac responses corrclale well with Hediger's (1955)
concept of flight and critical distance reactions in wild mammals
(see Figure 15). Upon this proximily-related schema of behavioral
and physiological reactions, those reactions of a sodalb.ed animal
reacting to conspedes or to man. if it is tame or domesticated, may
K- superimposetl.

Behavior, Development, & Psydtopathology of Cardiac Activity 217

Flight and critical distance reactions are then absent^ but phys-
ical contact still evokes bradycardia. This was studied in several
canids socialized to man, including domesticated dog, coyote, jack-
al, and wolf. . ,

Petting reHably evokes bradycardia provided it is accompanied
by an overt reaction of passive (friendly) submission. It is blocked
by active (greeting) submission or general excitement (see igure

16). .

Petting bradycardia results in a 10-30% reduction m heart rate,
similar decrease accompanied holding or picking up t e anima
Figure 16).

socia^rdt^thpa\"e%tL'L^^^^^^^^^

I Hold

Chewi bone I

Figure 16.

2i8

THE bog; its domestication and behavior

Alert Eating

t Groin Contact tOff

and muzzle seizure and piiming to the ground (see Figures 17 and
18), both important areas of social contact in canids (Fox, igTid).

Sustained bradycardia (but rates higher than during petting or
handling) was recorded during exploratory and investigatory activ-
ities (sniffing, looking at, or listening to novel stimuli) in all canids
(see Figures 16, 17, and 18).

This contrasts sharply with the transient bradycardia recorded
in man during orientation and attention (Lacey and Lacey, 1970). In
the vernacular, canids would be in a category of super-
"sharpeners," a perceptual-cognitive style in man associated with
great attention to details and to novel stimuli (or a change in the
familiar) with low habituation rates.

Heart rates during drinking were the same as or a little slower
than resting-alert, while during eating, tachycardia (excitement?)

Behavior, Development, & Psychopathology of Cardiac Activity 219

Figure i8.

J A-,. Pleasure response?) were frequently re-
or slight bradycardia (p Lji„ eating, a sudden, transient
corded. When '"“mi a freezing

bradycardia , 3, n,ore sustained following an actual

threat over the food ana was

fight (see Figure ’ 9 ^- . obsen-ations, the sustained low heart

From these compara ^ dominance (eliating passive

rates during ,‘e( in an unsodalired animal (produong

submission), aud ‘ gy „ common psychophysiologi-

freezing or tonic im

220

THE dog: its domestication and behavior

Sitting

Threat

Attack

Recovery

10 sec

Figure 19.

cal response. Some of these behavioral reactions in which sus-
tained bradycardia have been recorded are illustrated in Figure 20.
The possible evolutionary significance of this phenomenon is dis-
cussed subsequently.

222

THE dog: its domestication and behavior

General Discussion

These observations and the literature reviewed present a somewhat
paradoxical picture of the relationships between cardiac activity
and various behavioral states. Paradox is to be anticipated when
one is dealing with a complex neurohumoral homeostatic system
such as the autonomic nervous system. As Gellhom (1968) ably
demonstrates, paradoxical responses are often seen and can be
explained on the basis of sympathetic-parasympathetic balance or
ergotropic-trophotropic tuning in his terminology. Thus, an ap-
parently paradoxical response such as bradycardia can occur in
very different emotional states ranging from fear to pleasure.

An important factor in predetermining the type of autonomic
response may be the degree of sympathetic and parasympathetic
tone characteristic of the individual or species in question.

From the data presented, it may be hypothesized that wild
canids (of high social rank) have a Wgh sympathetic tone and a
high parasympathetic tone. Low-ranking, more fearful animals
generally have a lower sympathetic tone coupled with either a
weak parasympathetic tone or an excess of vagal inhibition either
of which may lead to maladaptive stress responses. In Pavlovian
terminology, the latter two typologies would be characterized re-
spectively as weak and imbalanced nervous typologies and the
former as strong and balanced. His fourth category, strong but
unbalanced, would be seen in those middle-ranking canids that
showed less adaptive mobility in inhibiting excessive sympathetic
or parasympathetic reactions but were able to adapt better to stress
conditions than the weak and unbalanced types (Kurtsin, 1968).

The effects of early handling stress would seem to be creating a
strong nervous typology with high sympathetic and parasym-
pathetic tone.

Domestication, from comparative studies of wild and tame
canids (and other species dted), may have selectively reduced
sympathetic lone rather than enhanced parasympathetic lone. It
would be naturally advantageous to reduce the fright-fight-flight
responses of animals in the process of domestication [and as dem-
onstrated by Bcylaev and Trut (1975) the adrenal-hypothalamus
axis responds in quantitatively different ways to stress in domesti-
cated subjects).

Behavior, Development, & Psychopathology of Cardiac Activity 223

Great individual differences are evident in dogs of different
breeds and temperaments, however. These differences do not con-
tradict the above hypothesis on the effects of domestication but
rather point to the original question of ergotropic-trophotropic bal-
ance and dynamics or mobiUty in the Pavlovian sense. Such indi-
vidual differences may stiU be manifest in the presence of reduced
sympathetic tone or lowered ergotropic tuning.

In his two classic papers on the effects of domestication and
selection on the behavior of the Norway rat, Richter (1954) states
that

Electrocardiographic records taken while the
strained shoi^ in the case of wild rats, marked slowing of the
heart — a definite bradycardia, and little or no c ange
domesticated rats. ThL it appears that, as compared with wild
ratS/ domesticated rats are much less vago onic.

Is this conclusion ^fU°^night'response

heart rates in wild and domesticated u -u-f of enforced

(and associated tachycardia) is not mobi e canids) then

restraint (analogous to passive inhibition m canids),
bradycardia might be a highly adaphve response,

Richter goes on to note that

Domesticated and wild '“^ffpr'^exposure m cold,

Woods, to the same form of stres ' jjj none

fighting, or exhaustion ^om svnmnungjn the wil^

of these forms of stress had any adrenals, whereas in

amount of ascorbic-add conten in , jevcl or

the domesticated rat it he Sorbic-add

else eliminated it altogether. In Hoses of ACTH.

content could be depressed on y y better

These results would thus indicate ‘ because of the more

able to withstand various forms ot stress-^
active adrenals.

224

THE dog: its domestication and behavior

passive inhibition (tonic immobility) in order to avoid either attack
by a predator or by a superior conspecific. Thus wild rats and wild
canids show bradycardia under forced restraint, and both wild and
socialized canids show bradycardia when being petted in associa-
tion with passive submission and friendly greeting. The latter be-
havior may have evolved, in the ethological sense (Chance, 1962),
as a means of inhibiting or cutting off a potential attack by a social
superior. This could account for the bradycardia occurring in two
apparently unrelated contexts (see Figure 20).

Increased parasympathetic activity (salivation, slowing of
heart rate, and increased peristalsis and secretion of digestive
juices) has long been recognized as a response in infant animals
and human neonates alike to maternal attention (Spitz,i949). This
parasympathetic (pleasure prindple) response in adult animals so-
cialized to man is readily evoked by human contact and petting, as
it is by one adult aiumal grooming a conspecific.

Absence of such stimulation in infancy may lead to impaired
growth (marasmus) and increased susceptibility to disease in both
human infants (Spitz, 1949) and dogs (Bleicher, personal com-
munication).

Thus, parasympathetic stimulation, an important aspect of
maternal care and neonatal physiology, may be linked with so-
cialization and potentiate the development of emotional attach-
ments. This has been elaborated upon in detail by Schneirla (1965).

In summary, therefore, contact with an animal socialized to
man evokes relaxation and bradycardia — the classic psychophysio-
logical response to petting or contact comfort. Contact with an
unsocialized animal, however, evokes bradycardia associated in-
stead with tension, freezing, or tonic immobility (see Figure 20).
This latter cardiac response is probably an adaptive homeo-
static mechanism to control for sympathetic hyperarousal. A
comparable compensatory inhibition was recorded in a canid im-
mediately after a fight with a conspecific (see Figure 19). Richter's
(1954) conclusion that domestication reduces vagotonia therefore
may be valid. High vagotonia would be an adaptive response to
sympathetic overarousal which would occur in association with a
wild animal's initial attempts to escape from restraint. Data on
unsodalized canids support this view; marked tachycardia when
approached but with escape blocked, bradycardia develops and is

Behavior, Development, & Psychopathology of Cardiac Activity 225

enhanced by physical contact (i.e., transition from freezing to pas-
sive submission). With a lower sympathetic tone in a tame or
domesticated animal, restraint or human contact would not
mobilize a compensatory homeostatic vagal inhibition.

But based on Richter's rather superficial data, vagotonia ap-
pears to be more a function of context than of domestication per se,
since human contact evokes bradycardia in a socialize
enforced restraint evokes bradycardia in an unsoaa e appe

wolf (M W. Fox personal observation). in

Domestication has more likely influence e
which vagotonic reactions occur and has lowere threshold

and modified the adrenal-hypothalamic stress reac ° . ^

(a point demonstrated by Beylaev and Tmt

docility over 14 generations). Consequent y, , "phasic"

thetic tone relative to wild species, the ^^P . be

vagal inhibition (rather than "tonic Richter's con-

less in a domesticated animal. ® „ beti^een the adrenal-

elusions helps clarify the interrelations P „ jystem and

hypothalamic sy“'P^*®‘'^'P®” 2 ^rs%drome (Cannon, 1956) in
would account for the sudden death ^ ^ ,g j animals which

wild animals under restraint and in „ which may trigger an

have an abnormaUy high ^^tta vaL^a^ark. ..

excessive phasic vagotoma leading Pavlovian "collision

This latter phenomenon, anatogous 10
of excitation and Inhibition (but ea * § ^ account for the

ganism rather than to a neurotic reac ' (be stress of

apparent absence of a plasma co so

restraint in low-ranking wolves. studies of temperament

Kurtsin (1968) has ^viewed exten^ ^

or nervous typologies of ^trated clearly how different

an conditioning procedures. He dem emotional or physical

temperament types react to ^bilitv to stress and disease

(e.g., total body ^“*“£,HalIv overreact, both auto-

is greater in those breeds that o® adaptive behavaoral an
nomically and behaviorally, an Future biomedical researc
physiological inhibitory n the role of the autonomic

might be advantageously ^‘’‘y^uwoarasympathetic intcrrelalion-
nervous system and sympat ^ stress resistance in og

ships in terms of disease susceptibility

226

THE dog: its domestication and behavior

of different breed and temperament. Again, as emphasized earlier,
handling stress in early life might be instigated to improve the
stress tolerance and disease resistance of susceptible breeds or in-
dividuals. As Kurtsin (1968) emphasizes, sympathetic overarousal
to stress, mobilizing high levels of adrenocortical steroids, may
interfere with immune responses and impair normal healing by
reducing the phagocytic activity of leukocytes. Hyporesponsive
individuals are no less susceptible either. Autonomic and emo-
tional (temperament) variables must be considered in understand-
ing individual susceptibility or resistance to disease (see also
Astrup, 1968). Given the wide variety of dog breeds, this could be
a fruitful area for future research.

Comparative studies of wild and domesticated cats
(Leyhausen, 1973) and canids (Fox, 1971b) have shown that wild
species are more active and responsive to novel stimuli, either
visual, auditory, or olfactory. Domestication, therefore, may have
affected the reticular arousal system and the response threshold of
exteroceptors, a process which may have been mediated quite sim-
ply by lowering sympathetic tone. Gellhom (1968) refers to this as
ergotropic tuning and makes the following observation, which
tends to support the above theory of the influence of domestication
(artificial selection for temperament-emotional reactivity) on au-
tonomic balance. A shift toward parasympathetic (trophotropic)
predominance is proposed.

Elimination of important afferent impulses tends to cause a
shift in trophotropic-ergotropic balance to the trophotropic
(parasympathetic) side. Thus, animals (cats) deprived of the
organs of smell, vision and hearing show an increased reactiv-
ity of the trophotropic system indicated by a lesser general ac-
tivity and an increase in the duration of sleep. Such animals
show also behavioral reversal. Nociceptive stimuli, which in
the normal cat induce ergotropic effects, exert trophotropic ac-
tions in cats which are blind, deaf and anosmic.

The drastically reduced sensory input is analogous to the
raised response threshold of exteroceptors proposed above. This
view differs from that of Richter (1954) who may have misinter-
preted the bradycardia of wild rodents as a general (tonic) indicator

Behavior, Development, & Psychopathology of Cardiac Activity 227

of vagotonia, whereas in fact it was an acute (phasic) trophotropic
reaction to stress. Its absence in domesticated rats could simply be
because they were essentially more docile and were protected from
such stress by having a lowered sympathetic tone and a greater
trophotropic tuning (or tonic rather than phasic vagotonia).

Murphree et al (1967) have reported changes in heart rate in
their two lines of stable and unstable (shy) pointers under different
conditions. As in the findings with wolves and German s ®

reported earlier in this book, the more outgoing sta e e a a
higher resting heart rate (approximately 120/min versus
the timid line). Contact with handler evoked a dear bradycarf a
only in the stable line, the decrease in heart rate bemg shgh'ly
greater when petted by a stranger. This latter resu '
by the authors, supports the hypothesis presen e 5„l,mis-

petting bradycardia may be related to passive ( ^ J strange
Sion, which in a sodalized dog, may be grea er

'’"'Most interesting, however, was the

nervous Une of pointers. There with an unfamil-

rate when the animals were being petted— neimer

iar nor a familiar person. . in order to monitor

All subjects were restrained on a ,, , , .uis variable

heart rates with a cardiotachometer. It is po ^ freely

may invalidate any comparisons betvwen e yjowever,

moving unsodalized canids , ‘Vupd m other contexts in

judging from the abnormal behaviors de behavior of our

the unstable line that they studied an ive are probably

subjects when human beings were iio j„gs The subjects

deaUng with two very nomal except for their

used in the present study were beha ■ ^ jo man early in

fear of man (because they had never een and lachycar-

life). Consequently, human presence evo bradycardia, and

dia, confinement triggered freezing* an jjj.jjciycardia — normal
physical contact resulted m from a genclicallv

physiological reactions. The unsla e P jp early life Their

selected line had received much approached

sustained low heart rates, absence o a rnntactcd, and catatonia

and of any further bradycardia when contac

228

THE DOG: ITS DOMESTICATION AND BEHAVIOR

suggest an excessive vagotonia or inadequate trophotropic balance.
Further studies have disclosed an increased incidence of atrioven-
tricular heart block, which together with low resting heart rates,
would suggest that these animals have an abnormally high vagal
tone (Newton et ah, 1970). This conclusion is further supported by
the recent finding (Lucas et al., 1974) that these excessively fearful
pointers lack the normal hippocampal theta rhythm (and had a
tendency to sleep more than normal dogs). (This accords well with
Gellhom's conclusions cited earlier.) The inability of these dogs to
habituate normally to novel stimuli was correlated with an absence
of normal hippocampal theta activity during wakefulness. It is
normally depressed (desynchronized) in the initial phase of an
orienting response but returns to theta rhythm as habituation takes
place. These dogs also manifested catatonia (inhibition of motor
function), which can be evoked in normal dogs by electrical stimu-
lation of the hippocampus.

A logical expectation would be that ergotropic rather than
trophotropic enhancement would occur during such orientation/
attention processing. Novel stimuli, however, usually evoke
bradycardia (as mentioned earlier in Fara and Catlett's study of the
guinea pig). Lacey (1967) has described a similar attention
bradycardia in man as a transient cardiac deceleration interpreted
as a "preparatory" autonomic response preparatory to subsequent
performance.

Graham and Ciifton (1966) have also reported transient
bradycardia in man as a component of the orienting response to
novel stimuli. Orbrist (1968) has shown that under these conditions
cardiac deceleration is accompanied by a decrease in respiratory
frequency and amplitude. These orienting- and attention-related
bradycardias thus differ from the fear bradycardia, which is of
much longer duration and is usually accompanied by polypnoea
(Hofer, 1970). Thus, the entrainment or influence of respiration on
heart rate is probably not significant. In petting bradycardia in so-
cialized dogs, muscular relaxation and decreased respiration are
associated with a decrease in heart rate. In a fearful unsocialized
canid, respiration may be increased but bradycardia is elicited fol-
lowing contact, and the animal is in a catatonic state of motor
inhibition (see details later).

Behavior, Development, & Psychopathology of Cardiac Activity 229

Conclusi9ns: The Phylogeny
of Submission —
Bradycardia and Affection

The following theory is an attempt to resolve the

between the bradycardia of fear and friendliness, n ac , ,

be no paradox. The bradycardia associated wit passive

submission may be a recent evolutionary deve opmen i

social vertebrates from the bradycardia of fear, rharacteris-

sive inhibition, freezing, catatonia, or tome ^

tic of more primitive vertebrates and inverte ra

Consider the modes of reacHon imSy

FUght or attack with effective antipred-

(playing dead), in ’"^riy pr^-speaeS' is wgh^y^^^^

ator adaptation. In such a behavioral sta , associated with

tained (as distinct from the transient bradycardias assoaat

the orienting response and attention). nhvloeeny, and in

It is often stated that ^nic im-

the context of the present theory, this m y different species

mobiUty is present in young mammals or r ^orc

(deer, fox) and tends to disappear wi _„rcists throughout life,
primitive vertebrates (rodents, reptiles), i P ^ gj by a sudden
In the former species, it occasional y tnnic immobilit)’ in an

stress, i.e., the shock of hysteria evolong tonic

adult dog (see Figure zi) animal hypnosis (Vbl-

This phenomenon has been bring about a rcsto-

gyesi, 1966) and to sleeplike conflict situations where

ration of homeostasis by de-arousa t actions preparatory to

displacement behaviors resemble ce gjjgd that these steep-

going to sleep [DeUus, ^967])- 1* ^ould b ^ by ek-c-

like behaviors, associated with bradyca .

trical stimulation of the carotid . ] displays, it 'S "'d ccc

In the evoluHon (phylog«^"y> context may occur m

ognized that some that are . por example, the e en

other species in a very is thought to be the ei o u

sive fear-grimace of primates and can.cts

230

THE dog: its domestication and behavior

Pjf ‘^'•sor of the submissive grin displayed in other spedes
^ J’® signa CFox, 1971). A similar evolutionary emandpa-

SecHo? of fear and contact-

If the bradycardia causes a change in behavior foUowing some

in ense noxious or fearful stimulus, i.e,, decreased activity and

Behavior, Development, & Psychopathology of Cardiac Activity 231

balandng of sympathetic overarousal, natural selection would
favor inactivity if it was an effective antipredator strategy.

Similarly, natural selection in social contexts, as where a
superior (alpha wolf or dog's master) evokes passivity and submis-
sive displays, could operate on these same primitive psychophysio-
logical (cardiovascular-CNS) processes. Chance (1962), for example,
has emphasized how passivity may cut off a possible attack y a

social superior in primates.

But this is not the end of these tentaHve correlahons. A passive
or actively submissive (greeting) dog or wolf will usua y ®
several infantile behaviors, such sodo-infanti e ac ons g
characteristic of more sodal, gregarious mammals (Fox,
regression to an infantile mode forces us to consi
parental care in infancy. In socially dependent a u , . j

a persistence into maturity of some aspects of e psy P j
ical dependency of infancy (discussed earUer, w^er^ parentaj be_
haviors evoke parasympathetic arousal an ^ theory of

arousal in accordance with Schneirla s [19 5] P

behavioral organization). parasympathetic

This initial dependency of the 1 of develop-

stimulation (contact comfort) leads to a ^ persistence of

ment toward sodal dependency j" of submission, greet-

analogous physiological changes in essentiaUy inner-

ing, and care-solicitng. In other w ' arousal and

directed homeostatic function of para X P immobility)

sympathetic de-arousal (as in the bra F jn higher verte-

of primitive vertebrates becomes o er psychophysiological

brates and instead occurs in sodal ofaffeclional

dependency upon others forms the basts/ ^ sodal contingency

bonds. Petting a tame snake or croco 1 reinforcement,

without reward, whereas a a dog t^fUral dependcnc)’ is ver-

The evolution of this psychop ys> ^jeprivalion of contact-
ified by the wealth of data on pot thrive and ma>

reinforcement in infant mammals. Z]in,vinc the loss of a mate
even die. Bereavement and — ^ic of many species also

or of a close companion in adult . ‘u^gjoiogical mechanism
demonstrate how pervasive this P^^ ^ other-directed depen

for social dependency is. Withou godal and leads a mon.
dency, the adult organism is relatively asoo

232

THE dog: its domestication and behavior

or less solitary life. Deprivation in more sodable species (dog,
rhesus monkey, and man) will impair the normal development of
dependency relationships. Such overly inner-directed individuals
have an extremely limited capacity to form normal social relation-
ships in maturity (asocial, autistic, or even sociopathic symptoms
may develop [Fox, 1974]).

In conclusion, therefore, it is theorized that the parasympathe-
tic nervous system (and cardiovascular CNS afferents) has evolved
from an initially endogenous homeostatic mechanism to an
exogenous, socially dependent mechanism is more gregarious and
sociable mammals. This "social (evokable) homeostasis" or etho-
stasis forms the basis for the development of affectional bonds be-
tween animals and between animal and man.*

Summary

Several studies of cardiac activity in wild and domesticated canids
in various controlled conditions were described. Since changes in
heart rate are indicative of sympathetic/parasympathetic activity,
monitoring of cardiac activity provides a sensitive, quantifiable
index of autonomic functioning and related emotional reactions
and overt behavior. Developmental changes in heart rate were
studied in several breeds of dog; with increasing age, between 2
and 3 weeks, a gradual slowing in rate was found. Treatment with
atropine demonstrated that this was due to the later development
of vagal inhibition.

Evidence was presented to support the hypothesis that heart
rate, along with other measures of emotionality and behavior, are

•In man, and possibly in other mammals, this socially linked and integrated
psychophysiological mechanism may be the basis for empathy and altruism. (Lacey
and Lacey, 1970, (or example, describe one experiment where immersion of the
hand into cold water evokes bradycardia; a person observing another doing this will
also show a transient bradycardia which is greater if he has recently immersed his
own hand into cold water.) Parasympathetic control (corticovisceral feedback) is an
essential component of meditation and certain Yogic practices; the physical and
psychological benefits from such "mind-body control" for man are now being ob-
jectively measured, so far with extremely promising results (Wallace, 1972)-

Behavior, Development, & Psychopathology of Cardiac Activity 233

valid indices of temperament. These factors, which as a multifacto-
rial constellation make up what is generally termed temperament,
were correlated with individual susceptibility or resistance to phys-
ical and psychological stress and disease.

To what degree such factors, and temperament per se, are
inborn or innately predetermined, was partially answered by
studying individual differences within litters. Environmenta in u
ences, especially stress early in life, were shown to have ® ^

cant effect upon temperament, including heart rate and ot er e
pendent measures such as plasma cortisol levels, dominance, an
exploratory behavior. . t t • u

Following the findings that those canids wit
heart rates (either innately or as a consequence 0 .

manipuIaKon) were superior animals in terms o ^

havior, learning ability, and resistance to stress, a g
over 100 puppies was conducted. The heart rate was “
several indict of temperament evaluation fo«nd ^o^

valuable but not exclusive prognostic measure of later
for training such dogs for miUtary work (guardmg and locating

concealed mines). . chnwn to evoke a

Using biotelemetry, contact and pethng reliably evoked
clear bradycardia in dogs. This phenomenon to occur

in dogs socialized to the handler. Bradycar unso-

also following physical contact or wit reactions oc-

cialized canids. The psychophysiolo^ mntexts was discussed
curring paradoxically in such very 7ntexte

in relation to parasympathetic-symp« u _ individual dif-
These observations, together wit ,7 stress experience,
ferences in cardiac activity and effects ° . Qn,ic tuning, elabo-

were integrated with a relevant . differences in temper-

rated by Gellhom, to account , ^an. The possibility

ament, reactivity, and stress suscepb ■ jncreased parasympa-
that domestication has contributed o differences m cardiac

thetic (vagal) tone was explored; and su.sceptibibty to

acHvity as an indicator of ^ for further study and

stress were emphasized to be a vabd subject
clinical evaluation.

XI

Domestication and

Man-Dog

Relationships

In the foregoing material, some of the ""^„7fiffu-

involved in domestication is discusred. O sp raisins pet

ture research, in both farm animal husbandry 'n « pel

animals, is the appUcabiUty of early ^

rearing methods are no less relevant m laboraton^ animal saence

^“'T^frlarch studies reported herein add mjre^U^-r^unde.

standing of the temporal “rgart^aHon o e , j^^rc re-
itance in wild and domestic “"ids and their nyP^^^^
search is needed on the ®“l’l*e mtenn i

vocalizations, although the study rep , ^ jog and the

upthequesrionastotheon^n^M^ „,her

possible effects of neoteiy an p^^^

behaviors. Domesticated dogs In subtle effects of domcsli-

able statement for the reversibib^ Hosdc dofExperience -vith
caHon and of the adaptability of ,he role of

hand-raised wild canids adds j ju chapter 2 , and the

sodo-ecological preadaptations . ,,icalion. This concluding
genetic or species

summary helps clarify and in eg • domestication and

into a more concise appraisal oiin
human relationships on canine

a.'s

THE DOG: ITS DOMESTICATION AND BEHAVIOR

236

Miscellaneous Aspects of
Domestication

It should be emphasized that movements and display postures are
more stable under domestication than other characteristics (e.g.,
display structures); thus, there may be little interference with
courtship, mating and other social behaviors. How an animal
communicates and displays may be less affected by domestication
than when and to whom it will display and when and with whom
it will Interact. Interactions between different species may affect the
amplitude and frequency of action patterns rather than alter the
patterns of behavior per se, although in some instances, provided
with the right contingencies of reinforcement, new behaviors
within the animal’s repertoire but not normally manifest may
emerge, e.g., canid mimicry of the human grin (Fox, 1971b) and the
acquisition of American deaf-dumb sign language in chimpanzees.

Evolutionary changes may be compressed and mimicked in
domestication, genetic mutation, selection and migration, and iso-
lation of populations under human influence occurring. Geo-
graphic and experiential factors (e.g., sexual imprinting) may en-
hance or reduce isolation in developing new hybrids and strains.
Scott (1968a) succinctly states for the dog that

The dog is . . . not in any large degree a conscious product of
human ingenuity. Rather it has evolved under the influence of
countless thousands of interactions with human masters. We
can therefore think of the dog as a species which, on domesti-
cation, entered a new habitat and underwent a process of
adaptive radiation similar to that of a wild species entering a
vacant ecological niche. It subsequently underwent further
modification and diversification as it became divided into small
local populations and selection pressure became relaxed in cer-
tain directions and increased in others under the influence of
the human social environment.

The domestic environment can also induce a new phenotypic
expression of the same (wild) genotype. Abnormal behavior and
genetic instabilities may rise, as in the regression to a wild temper-
ament, in overmagnification of a wild trait by a change in

Domestication and Man-Dog Relationships

threshold, or in disruption of temporal sequences such as sexual or
maternal behavior. Examples include, respectively, shyness, fear
of strangers, and neophobia which develops in some pups around
10-12 weeks of age, as in the wolf cub normally; hyperaggressivity
and proximity intolerance in certain breeds; aggressive dominance
disrupting mating and prey-stalking and killing appeanng m
bitches with pups. Repression of some natural behaviors may also
occur, e.g., normal regurgitation of food for pups is often absent in
purebred bitches and parental behavior incomplete or absent m
male dogs, since the latter, kept for generaHons without contact
with mother and pups, may show little interest in them and may
even compete with and kill them. Many questions remain
swered, notably why the apparent high mutagerac po en a o
phenodeviant forms in the dog in contrast to t e “ some

or is this mainly due to the dog having been rnginlv for

4000 to 6000 years longer and the cat paving , P ,
one role (to catch vermin) and not selective y , ^ e j
different roles as is the case in the various ^reed-c asses^of^dogs^
The question of the influence of coat co or o P influence
an intriguing topic, since selechon ^^oa example,

behavior and emotional rea'^^'^ity. Trumier
notes that black Alsatians are livehe an g tempera-

black is recessive, and some

ment and carry the produce black offspring,

black temperament are bred they P .. gj (hg possible
Keeler (19^5) and Guthrie (^97^ ha- J, fox in
significance of melanism in coIo p j-janee to social stress,

relation to temperament, ioral and physical data from

A multifactorial analysis of ^ graduate stu-

over 60 captive wolves was con ly^ he found a coirela-

dents, Randall Lockwood (1976). y^dult black color-phase

tion between behavior and 'and more active and pas-

wolves showed less temtonal ^nl^r-ohase wolves,
sive submission than brown or an^^ ,vhen a wild and domestic
Trumler refers to ' Hecks aw 1:0:50 distribution of re-

species are crossed, there ^^iT^hn^dribut rather "qualities came to
spective parent traits in the y ^ of the domestic animal,

light which were those of difficult to interpret in the

precisely what this statement implies

238

THE DOG I ITS DOMESTICATION AND BEHAVIOR

absence of further data, qualification of the concept of prototype,
and references to increased variance in Fj hybrids due to indepen-
dent assortment of Mendelian units. This latter phenomenon is
surely a key to increasing diversity and selection of phenodeviants
in early domestication. Trumler may be confirming this and Hecks'
law with what Konrad Lorenz (1970) demonstrated several years
ago in his studies of different species of duck and their hybrids. He
found that a hybrid of two species may show certain behaviors not
present in either parent species but present in another, possibly
more ancient or earlier evolved species. This is termed reactivation,
a phenomenological construct which may be interpreted in terms
of threshold and temporal sequencing of behavioral units, where
omission, reorganization, or repetition of various units may occur as
a consequence of either natural evolution (speciation) or domesti-
cation.

Some of the specific and potential changes attributable to
domestication in the dog are reviewed in this book. The changes
fall into a number of more general categories relevant to the effects
of domestication in other species as well.

First, the changes may be genetic or environmental in origin,
or they may be a consequence of more complex genotype-
environment interactions (Fox, 1970b). Second, it is important to
define the level of behavioral analysis from which conclusions and
inferences are drawn; for example, certain consummatory acts
(fbced action patterns) such as those associated with ingestive, ag-
gressive, and sexual behavior may remain unchanged, while then-
threshold, amplitude, and frequency may be increased or de-
creased. Also the temporal organization or sequencing of action
patterns may be modified (e.g., truncated at various junctions as in
prey-catching and killing in canids) as well as the type of stimulus
or release which evokes such behavior (e.g., a moving vehicle may
become substitute prey for the dog or a person may release sexual
responses in an imprinted bird). Generalizations, therefore, cannot
be made as to the direction of changes in behavior (see Figure 1).

Genetic influences are generally attributable to inbreeding, di-
rected selection, and genetic drift. The removal of natural selection
pressures and of such natural variables as seasonal changes in
light, temperature, and availability of food can evoke rapid
physiological changes (such as the multiple-estrus of wild felines in
captivity, which in the natural state are monoestrous). Structural

Domestication and Man-Dog Relationships

239

ONTOGCNY OF CANINS CTHOGHAM

changes, especially in display social marks) may ^

(horns, bumps, body ^tive breeding or more rapidly

brought about slowly ‘trough sel^ ^ vanabics.

by husbandry procedures sue

240

THE dog; its domestication and behavior

together with the social composition ot the captive animals (kept
together in natural groups, free-ranging, or confined in groups of
the same sex and age), can produce further changes in social be-
havior, growth rate, sexual behavior, etc. Thus, the method of
animal husbandry or mode of livestock management can greatly
influence the animal's phenotype and presumably, over time and
generations, the genotype as well.

In an interesting experiment with rats, Hughes (1973)*
evaluated three major environmental variables which may effect
docility, namely, cross-fostering (of wild offspring onto a domesti-
cated mother), rearing in an enriched environment (seminatural,
complex cage as distinct from the small barren cage of standardized
laboratory animal care), and early handling or gentling of the wild
offspring. As might be anticipated, only the latter procedure had
any significant effect in enhancing docility.

Thus in order to evaluate the effects of domestication, the
environment which the species is exposed to and is raised in must
be known and operationally defined in terms of its actual and
potential influences upon behavior, structure, and endocrine
functions. Since the domestic environment is often unnatural to an
extreme, abnormal yet adaptive changes in behavior, structure,
and function may develop through phenotypic change or genotypic
selection; atypical strains or genotypes may then emerge, quite
different from the original wild form (see Figure 2, Chapter 9).

Domestication may generally increase genetic variability and
consequently lead to a greater degree of phenotypic plasticity,
since the more unidirectional pressures of natural selection have
been eliminated to varying degrees. Some researchers also believe
that domestication causes a general degeneration, both physically
and psychologically. This may be true of certain systems or struc-
tures in some spedes, but as emphasized earlier, no single
generalization as to the direction of the effect of domestication can
be made. Degeneration, for example, may be extremely difficult to
distinguish from infantilism or neoteny (paedomorphosis).

A thorough analysis of domestication effects in any single
spedes must not only take into account the genetic, environmen-

•Hughes, C. W. (1973). Early Experimce in Domestiintion. Unpublished doctoral dis-
scrtiition. University of Missouri, Columbia, Missouri.

Domestication and Man-Dog Relationships

241

tal, and experiential variables discussed earlier, it must also en-
counter the limitations of comparative studies with a wild model
ancestor. The latter, adapted to captivity for study purposes, may
differ in many ways from one bom and raised in the wild, for
which it is preadapted. Similarly, the domesticated species may
differ from one that is raised under seminatural conditions or has
actually become feral. The conscious or coincidental changes
wrought by man's intervention must also be determined, for
example, rigorous selection for docility may increase proximity tol-
erance but may negatively influence fertility. Similarly, paradoidcal
selection may lead to behavioral problems, as in selecting Alsatians
to be highly sensitive, alert, and reactive; a number become
hyperactive, overreactive, and extremely timid, also selecting an
training a dog to be a protector or guard of person or
set up a conflict with the demands to be soaable and fnendly a
other times. Such coincidental or unidirected -changes and
paradoxical sequelae to consciously directed selechon
for specific utilitarian purposes must be iden e ® ^

complete understanding of the effects of domeshcation can be

gained.

Origin of the Dog

The origin of the domestic dog, *^P^esticated^e dog

though!: is generaUy held that man ^riaUy^

because of its value as a hunting p association between

(1965), cast doubt on this. His °f*“Xgo is a virtually

Australian abongines and its main value is to

untrainable canid, even when h g^^ement), to keep its

keep the camp clean of garbage (an gggasionally to act as a

human companions waim at rug r They are effective as

watchdog, giving warning ° ™ forests, a fact supported

cooperative hunters only m trop himtinc in a similar ram-

by the use of the Basenji in Je early stages of

forest environment. Presuma y 5 huntmg

domestication may not have invo v Inter, 'vith solec*

abilities except in suitable environments and >

242

THE dog: its domestication and behavior

hve breeding, did the dog come to fulfill a hunting role. Initially, it
would seem that the dog was a camp companion, follower, and
guard, and later its roles as hunter, draft animal, protector, and
herder of livestock were developed. Domestication of the dog took
place while man was still in the hunter-gatherer stage according to
Dr. Barbara Lawrence’ who has found remains of dog in North
America dated 8400 B.c., which indicates that the dog was the
earliest of man's domestic animals. The dog later became firmly
established in village farming communities, at which time, pre-
sumably, selective breeding for particular utilitarian functions was
undertaken.

Epstein (1971) has presented the most extensive and detailed
review of the origins and descent of the pariah and other types of
dog, with particular emphasis on those of Africa. He reaches the
conclusion, based upon extensive critical review of archaeological
reports, that the small southern asiatic wolf, C. lupus pallipes, is
most probably the dog's ancestor, discounting jackal ancestry on
the basis of chromosome number (jackal, 74; dog and wolf, 78).
However, Chiarelli (1975) reports 78 chromosomes for the coyote,
wolf, golden jackal. Cape hunting dog, and domestic dog; chromo-
some numbers are therefore of questionable significance.

Epstein (1971) concludes that

Initially the domesticated dog was of no economic use to man,
save for its doubtful role as a scavenger. Only gradually, and
in the course of its diffusion, it became a utility animal — an ob-
ject of sacrifice, of ceremonial or profane consumption, or em-
ployed as a guardian of the home and flocks, in the hunt, for
draught and for its wool, or merely tolerated as a scavenging
pariah. It may be assumed that it was only after realisation of
the usefulness of the first domesticated species that the idea and
practise of domestication were transferred to other species.

While this is a plausible hypothesis which I would support, it must
surely refute his notion (and the widely held view) of wolf ances-
try, since the latter would, because of its size and social behavior as
a pack animal, be an unlikely carud candidate for Camp/settlement
scavenger. Small contemporary subspecies of wolves in Southern

•Museum of Comparative Zoology, Harvard University.

Domestication and Man^Dog Relationships

243

Europe and Asia have adopted a more jackal-like mode of a solitary
scavenger because of shortage of available prey. It is improbable
that early man lived in regions where prey was so scarce, but rather
both the wolf pack and the human hunting-gathering group
shared a comparable niche rich in available game. The role of
scavenger would then be assumed by jackal- and dingo-like canids
in association with human campsites and following wolf-pack

hunts.

With the advent of agricultural settlements and increa^ng de-
struction of natural habitat, some wolves may have become
scavengers close to such settlements, while others move r ^
away into more remote uninhabited regions. The ormer wou
have been a threat to the Uvestock that were J

date and would have certainly been discourage . Jac a s, /

are smaUer in size and less powerful than wolves
been tolerated more since they would be bttle Yem^f “

people. In a parasitic/symbiotic mode compare e 0 domes-

jackals and feral pariah dogs, the primihve pro otyp j^ggegted
ticated dog was undoubtedly tolerated by early man as
by Epstein (197X)]- and this prototype ^^/dcateS

wolf-like. Therefore, the ’ (possibly dingo-Uke)

dog arose from a more primitive pro ^ man's interven-

which evolved from the wolf/jackal P
tion and prior to its initial ^(1068) noHon of the affin-

One must seriously queshon Sc ( 9 and pack

ity between dog arid wolf m " evidence to support the

formation (which, argues, is s g First,

view that the wolf is the ancestor o -vailability, distribution,
environmental influences, especia y fnrmation (Fox, 1975a)-
and type of prey affect sociability an^pa^^^^^ adopt^a

For example, in Italy, wolves tend ^ coyote. The

scavenging mode of ^^stence simto n,ay be

latter two species, normally living plentiful- Thus, envi-

found in packs and in areas where ^ there is a compara-

ronment influences sociability, an yj,i5 influence, keep-

ble picture complicated by man s m territoriality and has a
ing a dog on one's py°P^' free-roaming and ho/oeless

socifugal or dispersing effect, dominance hierarchy a

dogs in the same locale may establish

244

the dog: its domestication and behavior

hunt in a pack, provided food is plentiful. This was observed on a
recent field trip to Southwest India where a study of village pariah or
pie-dogs revealed this environmental influence on social behavior.

Studies of Pariah Dogs

Observations of home-owned free-roaming and feral pariah dogs
in India revealed the complexity of their social organization in the
typical rural village environment. Adopting and socializing one
adult female and following her interactions with neighboring dogs
added further insights in support of the following tentative
generalizations.

There are basically three types of pariah dog: (i) home-owned
and not inclined to roam more than a few hundred yards from
home, (2) home-owned but free-roaming, (3) ownerless and free-
roaming, with or without its own home-base territory, a den, or
shelter. Gregariousness increases from (i)-(3), attachment to man
having an intraspecific socifugal effect. In (1) the dog often man-
ifests relative dominance (Leyhausen, 1973) in that within its
home-base territory it is dominant over intruders. In (1) its
foragingihunting range may be restricted by two factors, namely, if
it is given food by its owner (which is rare) and if it is subordinate
to neighboring dogs. Foraging and hunting ranges overlap, and
consequently, where dogs meet away from their home bases, abso-
lute dominance is seen. While a dog may drive a neighbor off its
(owner's) property, it may be subordinate to the same dog on the
neutral territory of a shared foraging/hunting range. Thus, attach-
ment to man and place has a socifugal effect, while sharing of the
same food area leads to the more typical dominance-subordinate
hierarchy. Here, competition over available food concentrated in a
limited area increases conflict, and an absolute dominance hierar-
chy is established. Interestingly, though, this constitutes a loose
social unit (not a pack per se) and pariah dogs of types (1) and (2)
especially will temporarily pack together to hunt deer (an activity
encouraged by villagers) and to drive off a strange dog that enters
their shared range or collective territory.

Domestication and Man-Dog Relationships

245

Different relationships between the sexes, between dogs of
different ages, and females in heat and with pups also influence
the type of interaction between conspecifics of all three classes,
ranging from active and passive submission to indifference or overt

aggression.

Type (3) dogs may aggregate more frequently and form tem-
porary packs, sleeping, foraging, and hunting together, since they
lack the sodfugal influence. of having an owner and home-base
territory to defend. In larger villages, where several sources o 00
may be present (butcher and baker stalls), dogs may set up a tern
tory around a source and keep others away. Many sue fS® •'re
homeless and belong to type (3), but competition f

food source sets up a territorial situation ana ogous o ,

based dogs of type (i) and (2). In one village, t

pariah dog that was the most dominant dog m e <

having unchallenged freedom of movement an ^^ . ..

sources. In summary, the influence of j_(r£,ijm.es

more dogs with a home (shelter with or wi ou

sodfugal territoriaUty which is compUcated

dog. Lr compel. .1 copemle

defending their shared range or collec p,npe docs of all

dent free-roaming and feral dogs. With sh-^d ^ S - ^^^8
three types may be involved. by these socifugal

upon a social dominance hierarchy, neiehborinc dogs can

factors, but even so, this loose soaal "^‘Shbo^g g

form a temporary pack in a collective

range/territory against 5 bsequently, though, the pack

forest or jungle close to the village S^qu >; ^ f.u-,

dissolves into a loosely structured „»urn to. Also, ihe suppH

that some dogs have home terriloncs o deters pack fornu-

and distribution of their usual email numbers m order ti»

tion, since they must forage , optimi/o the utili/atum oJ

minimize competition and confl
such resources (see Figure z)-

246

THE dog: its domestication and behavior

Figure 2. Indian village pariah dogs showing a typically emaciated
bitch nursing pups and males forming a temporary pack around a
female in heat. Village dogs tend to establish territorial zones
around local food sources but in the absence of such may form
mixed nomadic packs.

alone. In the latter, where dogs share the same home range with
others, there is some form of sodal organization as a loose unit,
since they will band together and drive off strange dogs, as I also
observed in India. Presumably the interrelationships among
domesticated dogs with homes, homeless village-resident scaven-
gers, and more peripheral feral dogs are a complex interplay of
territoriality, competition/cooperation over food (garbage and wild
prey), with each other (and with wild species such as jackals), and
social dominance-subordinate relationships relative also to the
proximity of home base or natural den/shelter (i.e., the center of
each animal's territory). CTerritorial zones may overlap and several
dogs may share the same hunting/scavenging range). Sociability
will vary, therefore, not only in terms of the individual's tempera-
ment but also in relation to these sodo-ecological variables of space

Domestication and Man-Dog Relationships

247

(place) and availability and distribution of food. Since pariah dogs
may live singly or in packs, the flexibility in their social behavior
points more to environmental adaptations than to support of the
view of wolf ancestry as proposed by Scott (1968a). Menzel and
Menzel (1948) conclude that

The question of the pariah dog is among the most interesting
of zoological problems, particularly from the point of view o
racial history. We can be quite sure that an examinahon of the
history of the domestic dog in all its aspects would reveal that
there is scarcely a single facet that might not be il uminate y
research into the problem of the pariah dog. . . .

In spite of varying theories on the descent of the dog, mMt au

thors recognize h^o distinct sub-groups of the genus m the old

world which differ from each other in raaa is ““T

Northern and the Southern Dogs. The 'nike

s..*™ s,o«p. S>»d„ do.,

those in the Northern group) back to tne

kind of Canides. He writes as follows: ro than

Flood there existed a type o( Canides

the wolf, distributed in the same area “

number in the OTUsequent y ^pre^

Australian continent. This kind is d dinco."

rieties ... in the South it is represented by the dingo

.ueinc nf the domesticated

The pariah dog is a prototype Pf P ^naIoEOUS to those of
dog, Uving today in India ‘ communities. It shows a

early agricultural man in small vi g formation— which

high degree of social flexibility so ‘ conciusions pro

in no way supports or negates a wo ciudvinc the pariah dog.
or con can be drawn of wolf ancestiy fmm stuuj g ^
but rather, they manifest a degree o s unique to dog or ivolf

tive to the environment and which _.,nd jackal may form

Under certain conditions (Fox, ^ 9751 ’ may break up an

packs from a loose social unit, an w more solilaiy coyote

follow an adapHve social model typical of the m

and jackal. , , ,,rban and rural dogs in llw

Studies of free-roaming “bundance of food (prey m

United States similarly show tha " . V urban areas when- fo"
rural areas) dogs svill form packs, "

248

THE dog: its domestication and behavior

is less plentiful, packs are rare and a solitary mode of foraging
usually disrupts pack formation with a few exceptions. Dogs with
homes (and usually, therefore, a constant source of food) do not
generally roam far; they tend to be territorial rather than running in
neighborhood packs and the latter, when seen, are usually merely
temporary groups of males around a female in estrus as detailed in
Chapter 3. Thus, the contention that since dogs are highly social
like the pack-forming wolf, the wolf must be the ancestor of the
dog, is untenable.

Three other factors supporting wolf ancestry must also be
questioned. Scott (1968a) states that dog and wolf have very similar
behavior patterns and vocalizations. The latter is certainly not true;
the range and variety of howls recorded from a single wolf con-
trasts the limited range of howls in domesticated. dogs, as well as
other vocal characteristics discussed earlier.

Behavior patterns are very similar in dog and wolf, there being
a closer affinity between these two species than between dog and
coyote or jackal and other canids reclassified in terms of behavioral
rather than structural similarities (see Figure 1, Chapter 2). While
this evidence may lead to the conclusion that the dog is a domesti-
cated wolf, it equally implies that they may share a common ances-
try prior to domestication and that the dog was a dog before it was
domesticated. This idew is supported by the fact that behavior
patterns per se are influenced little by domestication; how an ani-
mal behaves is more or less phylogenetically fixed, but when and
to whom it behaves as well as the threshold and sequencing of
behavioral units is affected by domestication and early experience.
Structure is more influenced by domestication than behavior; a
bulldog, Irish wolfhound. Chihuahua, and dachshund, structur-
ally very different, share the same basic behavioral repertoire.

The change in structure which most archaeologists and
taxonomists recognize as a consequence of domestication is the
reduction in tooth size; even a large breed like the Irish wolfhound
has small teeth relative to its skull size and in comparison to the
wolf. This may, however, simply indicate that this breed type is a
giant mutant from a medium t5rpe having a tooth size more com-
parable to a large jackal or small wolf. So a small wolf has been
sought as the probable ancestor of the dog, and the Asiatic wolf,
Canis lupus pallipes, is now accepted by most taxonomists and canid

Domestication and Man-Dog Relationships

249

ethologists as the dog's ancestor. Yet there is no evidence of a
transitional form; all that excavations have revealed in the Middle
East to date are remains of a small wolf, large and small spedes of
jackal, and what is thought to be dog, with skull and jaw fragments
which overlap in measurements with the former two spedes and
so make identification difficult (Clutton-Brock, 1969). The contem-
porary native pariah dog is very similar structurally to these 8,000-
to io,ooo-year-old canid remains. Berry (1969)/ important
paper relative to this archaeological dilemma, points out that
natural environmental changes can mimic changes attributed to
domestication, e.g., a reduction in tooth and jaw size may be ims
identified in the case of the wolf as being due to domesticahon.
Pre-Pleistocene wolves were of much larger build in t e e

East than the existing subspecies today.

There is a cline of blotched tabby in European
to the coat pattern thought to be an exclusive trait
Berry concludes that

wild cats similar
of domestic cats.

even for a single species, it may be extremely
down criteria to distinguish between domesdca
domesticated forms . . . there is no reason to beue
domestication per se will alter the phenotype or

part of the genotype. . ,

° j r r f has been jdcn*

Further review of the earliest remains o p c. as the

dfled as C. familiaris gives dates of (i975)

earliest date of domestication. Interesting y. jineo by 20,000
ports that the Australian aborigines prece e dated around

y®3rs, since the first signs of dingo ,„d dog alike re-

10,000-8,000 B.C. Origin of dingo and domestic
main an enigma archaeologically. v,nlds that when it was

General (but unfounded) consensus denied

first domesticated about 10,000 ,,, hwus patUV^)

mm a wolf, possibly the Asiatic wo ( jj-po-like canid "as
*aeological records, however, show that a (Macinuwh.

'Widespread throughout Europe in the S b pn-

iq7ciu:__ ? .. .u-. dnmeshc dog ' .

"mespread throughout Europe m o- - 5 denveo t

■975). It is my conTention that the df ddfen;

anly from this prototype ^'’”’1^ j -vith mdigcdcds

place.; ..„c crossbred w'f ‘ m.ile.-nu

P sees and times this canid was guch

° produce some of the more wolfish bre

250

THE dog: its domestication and behavior

and the husky. Although the dingo in Australia is somewhat like
the coyote in its social behavior, being relatively solitary, living in
pairs, and not forming packs, it is the ecology that has determined
this kind of social behavior. If prey were more abundant, pack
formation might be possible. Today, the domestic dog is capable of
reverting to the wild (becoming feral) and surviving well, either
singly, as a scavenger or opportunistic hunter, or combining with
other dogs and forming a pack. The pack instinct of the domestic
dog and its responsiveness toward a pack leader may well be the
basis for its integration into the human family. In other words, a
transference of the pack relationship to a family allegiance occurs,
and this natural instinct greatly facilitates domestication and train-
ing. In fact, if socialization is delayed, the dog is virtually untraina-
ble.

Behavioral Effects of
Domestication

Further behavioral evidence of wolf ancestry from Scott (1973) must
also be chaUenged . He suggests tt\at the muizle-bite greeting ritual
of the wolf (detailed by Fox, 1971b as a spcdes-typical action) was
wlcctively eliminated in domesticating the dog since it is frighten-
ing to man. Having kept wolves myself and knowing others who
h.ive raised them and who arc familiar with their behavior, there is
no anxiety around such interaction. Also some northern dog
bmeds which probably do have recent rvolf origin (malemute and
husky) show this muzzle-biting behavior and breeders have no
concern or anxiety about this trail.

Domestication and Man-Dog Relationships

251

surely recognize a socialized canid independent of its tail! This
attractive conclusion is unwarranted however. It is not so much the
structure and carriage of the tail that is signihcant (even though the
Spitz-type dogs always carry it high) but the differences in tail
displays in dog and wolf. If it is accepted that structure (cur y or
docked tail) changes more under the influence of domes ca on
than behavior (tail postures in communication), then t e wo an
cestry of the dog seems even more remote. For examp ^
threat posture of the wolf, the tail is held verhca y
angle of 75°-85° from the horizontal (of the back), w ® ^

it is curved as an arch over the back, even f j

eastern pariah dogs) that carry the tail low in e sa p
the wolf when at rest (see Figure 3). mat ourports

Having negated aU “TndSsty ^

wolf ancestry to the domesticated the dog

supporting archaeological f ^J^ted his/her specu-

remalns unsolved. Lorenz (1975)- who ^ j},^

lationof a dual wolf-jackal dog ancestry, ^ as with the case

general consensus in favor of wolf ance^, g f^^enzel (1948)/
for the jackal, with no clear evidence. detail, point out

who studied contemporary pariah dogs 1 _ ^ ^e

the dilemma that one cannot today 1 independent of

domestic since the latter may become e . ] coyote, but
man. Dogs may also hybridize wit wo ' ^d usually

presumably ethological and ecoli^ca a q

prevent this. The point is that , panics which, like a wild

familiaris in the wild today other than e _ p£ phenotypic

of phenotypic uni-

species, will eventually breed to a g

formity after a few generations. j pif jj the supra

Aiiother dear cUstinction between dog en

caudal tail gland (Fox, igp^)- f ,he taU, is active m the

stripe of dark hair on the upper surfare ^t^^ , Furthermore, if

wolf, no odor can usually be detec e ^ jp me domestic 0^

there is a distinctive hair mark over jjy triangular as

the mark (of darker or thicker m (he wolf- This is s

coyote and jackal and not a thin stnpe

human nose, distinctively ■ Hogs'

’When an odor can be detected, it is,
contrast to the sweet-mown hay odor o

Figure 3. Dislinclivc trails of wotf and dingo, (a) Vertical tail in
dominant uvif (in center); (W arched tail in dominant dingos—all
asserting high rank; (c) and (d) rank conflict between dingos on in-
Iroduclion, including aggressive clasping and (inhibited) cheek-
oriented Hte; (e) in same period together wolves form an integrated
pack in contrast to more individualistic dingos: (P dominance-
submission in F. coyote X dogs; note doglike arched tail position.

Domestication and Man-Dog Relationships

253

strong evidence against recent wolf ancestry in the dog. (See Table
I for a general overview of wolf/dog differences.)

Comparative studies on the behavior and development of wild
canids has enabled us to identify a number of changes, whic in
the domestic dog may be attributed to the effects of domes ca on.
For example, the wolf reaches sexual maturity at 2 years ot age,
while other wild canids attain full sexual matunty at i year of age
In the wolf, this delayed sexual maturaHon might have important
social consequences, in that the young will stay amn" ° sexual
year to help adults raise another litter. In Domestica-

maturity is reached at a much earlier age (6-9 mo • , ^

tion, therefore, has accelerated the opment of ^ndocnne

glands associated with reproduction. Rela e °
full sexual activity in the wild canids is the nmerg . , in

dominance relationships, and defense ° time base at

the domestic dog, these behaviors appe central emo-

12-18 months of age. This suggests Uj ^jp^i

tional aspects related to full sexual ma u ty ,j^c>ueh domestica-
the endocrine aspects of reproductive behawor ,,,„al

tion. For example, by 1 year of age a eoy against in-
maturity and at the same male rivalry fights. In

truders and wiU engage in dominan attained much ear-

the domestic dog full reproductive po aeuression toward in-

lier in life, while problems assoaa e develop until

truders and dominance fights wit t e n sperm seasonally,

a later age. The males of wUd “™.^/P™ ,ntly fertile. Whereas a
while in the domestic dog the male 1 ^ the dog

wolf may show specific mate Pf® ^ 'jij make selective bree -
may have been eliminated, since jj ( sk if the stud or bite

ing under the choice of the owiier a ^ff «
refused to mate with a selected an beagles, and o

Comparative studies of identical conditions,

domesticated breeds, all raise {jg dogs show nme

shown that on separation the o observations

distress. It can be*^ concluded ^^pendent than the ».ld

domestic dog is basically mu the easier it is to train

species. Theireater thedependen^>- ex.rcm

animal. Hand-raised wolves and 7 potonousl)

able and friendly toward the another consequence

train. It may be proposed, there

Tnble I.

SOME EFITCIS OF DOMESTICATION: COMPARISONS OF WOLF AND DOC

o O •

= «}

:£ X :c

TJ

C

c c

fli ra

e s

ffl 6

.E o

1£3

^ o
5 ^
fS -S
1 g-ii
■« y 3
^ w 6

5 g 3

'« trt C

5 X

X (0

•?

■c _

"1

•si

o IS 4^ 01 o

S J c a. i

E g 3( E E

O q X o o

o

2

c

*c)

,S)

ts

•X

t5

OI

p

'o

c

c

•9

f7

•c

2

I

o

5.

S'

1 1

*2

2

X «

■a

1 .1

c

o

r?

1

f- 3

O

•Si

d :S

eo

.iggressivc hip-slam common rare except in wolflike breeds

vertical tail in threat* usual display in dominant wolf absent — tail arched in threat

T.iblc I. (nmliiiunl)

S . jj

SPS -3

« S S
S “ .5
>■ E S

ll’

I '<3

1 .p ni 5

« 5. a. 5 .5 S

■£ (fi ^ «

S « C C

o OJ 5 .53 *? ?

g X iS - O

CT- o c a br

so £ u o >

.2 <D 41

^•e1

c

E ^

? Q

cs| O W *0

bb

.< o

5 O,

S -5

. s'tS

ti>s ^

g=s ^

C

.B O Xi ^

•a

«0

2

^ *43 O

u - ‘C c

3 C 5 -

M > .W

S -9 <5 o

(B a» Cft

*« So TO

c S

“ ■§

CO £■

O) •&

•i o §

g '0 T3
^ C C
^ — TO

TO oi .a

E I:

TO ® C“

E S «

Q. a» i.

cw.g

TO >
T3 >

« c

CO "to Cli
.a C 4)

^ i K

*d

CO > w (y
3 C ‘43 ^ 3
CO ii cB E -O

U**'— >

•2 g

•S'S

3 C

CO 0*

3 -O

a -5 -3 -5

eo

o • yl"

2 ^ ^ a. c

•s 2 5 4( t< 12

I E E E

tfi

E w r£ -

a 0.8

•c —

& c

a> S

S a.

> -3 E

ef.

•Wolf-<log differences may imply separate ancestry rather than effects of domestica-
tion per sc.

Domestication and Man-Dog Relationships

257

domestication has been to make the domestic dog extremely de-
pendent. The degree of dependency varies from breed to breed.
For example, malemutes and huskies are aloof and are much more
difficult to train than the more dependent and willing terriers and
toy breeds. In the toy breeds, there is not only a greMer
dency but also a neotenization or infantilism, whereby t e P
structure of the adult animal more resembles an infant, le m
with the relative immaturity of the structure is the p^e ua °
infantile behaviors into adulthood. How much 0 t is
and how much is related to the way in which t ^ .

animal is open to question. The capacity to re am
haviors into maturity must be genetic.* Cornpara infantile

the wild canids reveal that the wolf has const era ^

behaviors (which are called derived infanti e ac 0
compared to other wild species. integration and

dency to foUow supenor animals, help njiveness to a

coordination of pack activity, . j ,i,e domestic dog.

leader or parent figure is no less evi Jemonstrable dcpen-
However, because of the greater de^ possibility for de-

dency in the many breeds of domestic S' time, it

veloping emotion^ disorders is ewance , . ^jy dependent

isamorefulfiUingpetforthosewhoneedanextrem y

relationship.

Man-Dog
Relationships and
Human Needs^

voung people

Human relationships are now „ children for a

having any children or are delaying citizens, retired
of years after their marriage, en

are not
numlvr
(x-ople.

Domestication and Man-Dog Relationships

259

parent, then it should not be a surprise that both dog and child
under certain conditions may develop analogous or homologous
behavior disorders. These can range (Fox, 1968, Brunner, 1968) from
psychogenic epilepsy to asthma-like conditions, compulsive eating,
sympathy lameness, hypermotility of the intestines with hemorrha-
gic gastroenteritis, possibly ulcerative colitis, not to mention sibling
rivalry, extreme jealousy, aggression, depression, and refusal to
eat food (anorexia nervosa).

Types of Relationships

The following is an arbitrary classificaHon of various types of rela-
tioLCs which may be established between the owner and the
dog (fL, 1976c): object-oriented;

de?efdence,^,^trV^^^^^^

guide for a bUnd person, or is used foxhound The use

driving cattle, or for sport, such ^ increase today. People
of the dog as a guard is paranoid about crime and vio-

Uving in suburbia are opberman pinscher or German

lence and will buy a dog such More recently dogs

shepherd that they will have clinical psychologists

have been used as camne the patient (Corson ft

utilizing the dog as a therapeutic bndge witn t F

»l-, 1975)-

Some Consequences of
Pet-Owner Relationships

It is the
can be the

h^ncnnn doc and owner that
close s^biotic Emotional and psychosom.v

he foundation fora numoeroi

258

THE dog; its domestication and behavior

widows, and widowers whose offspring might now be living many
miles away have a need for a close companion, be it cat, dog, fish,
or parakeet. It is not only the patterns of human social life that have
changed over the past 20 or 30 years, but also human needs as well.
The need for companionship, for example, is exaggerated when
families are separated and when people are lonely and alienated in
a depersonalizing urban environment. The more dependent the
pet is for some owners, the more fulfilling it is as a companion or as
a child substitute. It is this dependency, however, that opens the
doors to a number of psychosomatic and psychogenic emotional
disorders, some of which are analogous to those described by child
psychiatrists. I believe, therefore, that the incidence of some of
these disorders will increase and that the veterinarian in small
animal practice will have to be on the lookout for such disorders in
the near future, if not today.

Critics might say that to attribute a dog with humanlike emo-
tions and needs is to be unscientific and anthropomorphic. Re-
search has shown, however, that the developing brain of the dog,
its unfolding pattern of socialization and other critical and sensitive
periods during development are very similar, and sometimes iden-
tical, to the same phenomena recognized in the human infant,
although they develop on a different time base (Fox, 1971a). The dog
has basically the same limbic or emotional structures capable of
generating specific feelings or affects reflected in overt emotional
reactions and also in changes in sympathetic and parasympathetic
activity which are linked with psychosomatic and emotional disor-
ders. Add to this common neural substrate shared by dog and-
human infant the important variable of attachment which is a con-
sequence of socialization, as between dog and owner and child and

Domestication and Man-Dog Relationships

259

parent, then it should not be a surprise that both dog and child
under certain conditions may develop analogous or homologous
behavior disorders. These can range (Fox, 1968, Brunner, 1968) from
psychogenic epilepsy to asthma-like conditions, compulsive eating,
sympathy lameness, hypermotiUty of the intesHnes with hemorrha-
gic gastroenteritis, possibly ulcerative colitis, not to mention sibling
rivalry, extreme jealousy, aggression, depression, and refusal to
eat food (anorexia nervosa).

Types of Relationships

The foUowing is an arbitrary classificaHon of ° '

tionships which may be esfabUshed °

dog (Fox, 1976c): The most

dependence; transp«sonal re^ d ^ ,^^.^_^^P

general one is simply neea aepenuti ^ j

Irai. .h. «,d.l in ''“rr.t 3 ronr.(rSS.n

Another relationship with ^ dog P^J^ j ^ guard, as a

working relationship where the og i P ^ herding sheep,
guide forabUnd person, orisused orwork^^^^^^^^^^
driving cattle, or for sport, such increase today. People

of the dog as a guard is paranoid about crime and vio-

Uving in suburbia are ^ Doberman pinscher or German

lence and wiU buy a dog sue . More recently dogs

shepherd that they will ‘ ‘ 15,5 by clinical psychologists

have been used as canine co- P patient (Corson el

utilizing the dog as a therapeutic bndge with P

at., 1975).

Some Consequences of
Pet-Owner Relationships

It is the close symbiotic .md psydi>«.'ma-

can be the foundation fora num

26o

THE dog: its domestication and behavior

tic disorders. For example, the dog that is overindulged and is
raised literally as a child substitute may develop a variety of be-
havioral abnormalities when its relationship with the owner is
threatened — as by the birth of a child, by the introduction of
another pet, cat, or dog into the household, or by the arrival of
house guests. Separation from the owner due to the owner being
sick or the dog being boarded when the owner goes on vacation
can similarly trigger behavioral pathologies. These include unpre-
dictable aggression, depression, anorexia nervosa, hivelike reac-
tions and pruritis, excessive eating, sympathy lameness, convul-
sions, asthma-like conditions, cardiospasms, vomiting, and intes-
tinal disorders, including hemorrhagic enteritis. The overindulged
dog may also be underdisdplined and when it reaches full sexual
maturity it will behave like a socially maladjusted "canine delin-
quent." It may effectively win the dominance fight with its owner,
and it may become the overlord of the household. Such socially
maladjusted dogs are extremely difficult to handle on their own
territory and can be no less difficult to handle in the hospital. It is
important for the veterinary surgeon to establish his dominance
over such a dog, and it might be discretely done in the absence of
the owner.

In making a diagnosis, it is important to look into the family
background of the pet, and to be aware of the dynamics in the
household and to be alert to any recent changes within the home
environment. Some emotional disorders may disappear spontane-
ously when the animal is hospitalized and is removed from such
aggravating circumstances. A careful differential diagnosis must of
course be made and the possibility of allergic, organic, and other
infectious causes must be considered. The real crux of the problem
is that the close symbiotic relationship can be the genesis of a
number of dependency disorders in the dog which can be ex-
pressed behavioraliy or psychologically and somatically.

Domestication and Man-Dog Relationships

261

been neotenized nor have they been made much more dependent
than their rvild counterpart. The kind of person who will prefer a cat
to a dog as a pet is perhaps less Ukely to need to indulge such an ani-
mal (Fox, 1974). I feel that many people who need a dog are depen-
dent, they tend to be other-directed, and they gam considerable
emoHonal saHsfaction from having a depndent compamon m theu
Uves. However, more independent,
keep a cat in preference to a dog simply ‘ ®
and its less demanding atdtude. The breed
owns may be a projecHon of deeper needs and
insecure or parLoid person may want a J

Another person who is attempting to ve up galuja It is

grace and agiUty may keep an A^S^an hound or a ^
primarily because of these reasons that the pet otten

owner-it is something '""^^XTexperimental research and
Against a background ^ f beginning to under-

a scattering of clinical case luston ' ^ socialization influ-

stand more^ompletely how domeshcariona^^so^

ence the behavior of man s close future relationships be-

awareness will, I hope, not only improve
tween pet and owner but also the relationsmp
beings in general.

Postscript

Some may demean domesticated animals as being degenerate or in-
ferior forms of their wild ancestors or counterparts. Others may
see them as merely utilitarian "tools/' man-made to serve hu-
manity, in order to satisfy and gratify our many and diverse needs.
Yet do we fully understand our enormous obligation and debt to
them, which is ethically far greater perhaps than our debt to wild
forms? While the latter may be in our trust and we their stewards,
the former are our own creations. Being so, what kind of creator are
we, and are we to become? Our debt to them is unmeasureable, for
we have learned and are still learning from them to become more
fully human: responsible and compassionate. We can learn through
them in countless ways about nature and about our own nature as
well. If we knew everything there is to know about a hair on a dog's
back, we would know everything about the entire world, since
everything is interrelated in evolutionary time and global space.

As we gain wisdom, compassion and empathy through them,
and learn from our mistakes as they suffer for us, we may more
surely become the creative and responsible stewards and co-
creators of and for all life.

262

Appendix I

* » j 1 hr of Stimulation daily, comprising
The handUng procedures consisted soundproof box with a flashing

10 min photic stimulation in a lignt -—^fplv i sec frequency, lo nun
light stimulus of o.i6 intensity and . anteroposterior and 5 min

labyrinthine stimulation consisting o ^ 5^^ through an excur-

bitemporal tUting at an Auditory stimulaflon, 2 nun each at

sion of 45° from the horizontal, 10 of 1 o v and duration of J.o msec.
1, 10, 10= and 10’ cycles/sec at an intensity 01 ^ jn a
This was followed by 1 mm exposure m handling, dunng

centrifugal rotator at approximate X 45 ty ' the Magnus, rool-

which time a series of reflexes were reflexes (see Fox, ^9^5

ing, righting, geotaxic, pain, P®" reflex development could be

1966). By eliciting these reflexes, , groups. The subjects were

assessed and compared in ir^erLd for 15 sec, dunng

then placed in a water bath at during the first 3 wee ■

which time they would swim: “"wel and groomed wth a soft

Subjects were then rubbed dty with a ha^^j^'J'^^eous stimulation «ith an

brush for 10 min and received z mm handUng peno w

air jet (6o°F). From 3 weeks of a^ operator. After this bandl g

increased to include 10 min P'^X Control subjects were kept

period, they were returned to mother, hawng fl

under typical rearing eonditions wi* y elcaning

scheduled human contact of ‘«^“;^“‘/„ere recorded at '

subjects were weighed and b®® . of the invcstiga o _

while the pup was lying quietly in *e h= gjobul.n and piperazine

B weaned at 4 weeks of age, ^

were '

264

THE dog: its domestication and behavior

anthelmintic and were reared singly in metabolism cages in the animal
house environment.

At 5 weeks of age, the subjects were tested singly in a behavior arena
equipped with one-way windows to enable the experimenter to observe
the animals without being seen. The arena contained cloth bedding from
the mother of the pups and, in another corner, a brightly colored child's
toy. The reactions of the subjects were observed for 5 min when they were
first placed in the arena, and the observations continued for 5 min more
after these objects had been removed. The subjects were then replaced for
5 min more of observation. The observers with stopwatches indepen-
dently recorded the duration of certain activities of each pup throughout
each of the three 5-inin observation periods, and time scores were then
averaged for each group. The activities observed and time-scored were as
follows:

1. Specific Interaction with Stimulus. Duration of interaction with either
cloth or toy, including approach, play, chewing, licking, carrying, and
lying beside or running around the object, was recorded.

2. Nonspedfic Exploratory Activity. The time spent exploring the arena
was recorded, including sniffing and Ucking walls and floor and jumping
up at walls, looking up at walls (visual), and attention to extraneous noises
(air conditioner turned on as sound blanket). During this activity period,
pup never approached cloth or toy.

3. Random Activity. The time spent sitting or pacing the arena without
any overt reaction to cloth or toy or attention toward walls, floor, or ex-
traneous noises was recorded.

4. Distress Vocalization. As a level of emotional arousal, the number of
distressful yelps and the duration of distress vocalization was recorded by
one observer only while the other observer noted what else the pup was
doing (random or nonspedfic exploratory activities). There was a high
correlation between distress vocalization and nonspecific exploratory ac-
tivity.

After this 15-min testing period, the subject's approach to a passive
observer in the arena was determined, and then approach and following
response were assessed while the observer walked around the arena. The
ability of the pup to negotiate a simple wire-mesh barrier placed between
him and the observer was next used to assess detour behavior. Four trials
were allowed, and if the subject was able to come around the barrier to the
observer one end was blocked; if there was an end preference, the preferred
side was blocked first. The time taken to pass around the barrier and the
number of trials required were recorded. Finally, the social interaction of
these differentially reared subjects was observed when they were placed
together in the arena, which still contained the cloth and toy. After 5 min

Appendix I

265

of observation, the experimenter entered the arena and observed the ef-
fects of the presence of a human on the group behavior of the pups. After
these behavior observations, EEG recordings were taken on a Grass
6-channel recorder or on an Offner 8-channel type R dynograph, EEG
recordings were also taken and evoked responses to visual and auditoiy
stimuli of various frequencies were recorded on a computer of average
transients (CAT 400 B) and monitored on an osdlloscope via a Grass
6-channel EEG recorder. Recordings were taken while the subjects were
lying quietly awake and also while they were asleep in a darkened room;
they were retained in a copper-gauze box lined with foam rubber. Selected
subjects were then killed and several organs were dissected in the cold
room at 37“F. The brain was dissected for histological examination, and
the adrenal glands were dissected and prepared for epinephrine and
norepinephrine analysis. Some tissues were also prepared for lipid
analysis.

Appendix II

269

interval. (9) Finally the heart rate change following a 5-sec duration bell (75
dB) placed outside the plywood arena next to the comer where the cub
was lying is recorded.

STRESS STUDY

This study is based on the notion that confinement in a holding cage
would be stressful and that cubs might react to such stress differently
because of individual differences in emotional reactivity (which correlate
with sodal rank). Blood samples were taken at arbitrarily selKted
for plasma corticosteroid analysis with the anticipation t at in ^
differences in the intensity and duration of the stress . *

found. Six cubs from the two litters, selected on the basis o 1°

mnk scores, were each placed in a a x aVr X aVr ft
had been cleaned with Pinesol and lined with hay. e ^
placed in a dimly Ut and relatively soundproof room OnV/'™

used for each run, 4 « of „ ‘"f.j^^TtresTwas kept to a minimum
and 60 min after cordinement. Handbng s accessible and

while these samples were being taken, The hair over each foreleg

the entire procedure taking no more ^ the EKG study in

had been shaved the previous day Twenty-four hours after

order to minimize additional handlmg , .ccBM andRcBrd of Litter
these samples had been taken, four of the cubs (ScB d
I and ScBld and BIUc 2 of Utter II) were '"icc^ at the same

the time of confinement, and blood samp es k, attempt to elicit
intervals. The rationale behind this latter ‘ccatment^^as
a maximal plasma corticosteroid response, ihreshold and re-

the earlier samples, '"f’“^°"and duration might be disclosed

sponse magnitude as well as latency

268

THE dog: rrs domestication and behavior

min duration were run on alternate days for 8 days. As reported earlier
(Fox, 1972a), this test is of value in identifying the most dominant and
subordinate cubs. Middle-ranking cubs of Utter I, which showed greater
proximity tolerance and "shifting" or relative dominance (in that one
member of a pair appears to be dominant while in possession of food),
were impossible to rank in a linear hierarchy. This test was most valuable
in identifying the highest and lowest ranking cubs for later physiological
tests. For this dominance test, and for the evaluation of fear reactions and
exploratory behavior when presented with a novel stimulus, the same
ratings as detailed earlier (Fox, 1972a) were used.

NOVEL STIMULUS

Using the same horizontal stimulus as developed earlier (a 30 cm x 2 m board
with a 30-cm black square centered by a lo-cm white cross. Fox, 1972), the
latency of response or emergence time from the nest box was recorded for
each cub.

ULART RATL MEASURES

270

THE dog: its domestication and behavior

week for approximately 15 min for making growth measurements. They
were therefore essentially unsodalizcd but used to handling, during
which they would manifest passive submission, occasional defensive ag-
gression, and would attempt to escape whenever possible. All cubs
showed the flight reaction when approached, and only the two highest
ranking cubs would approach a passive person in the home cage to within
a distance of 4-6 ft; none made contact or showed active submissive greet-
ing. All cubs had received the same handling and no preferential treat-
ment prior to the time of testing.

Appendix III

273

wooden guillotine door was a start-box 26 in. square and 18 in. high. The
connecting door was 19 in. wide and 20 in. high. The start-box had a
hinged top which was fastened down by a screen door-type latch. The
arena was lighted by four 400-watt fluorescent tubes so as to light the
center square with 145 foot candles. The start-box was not lighted. When
the door to the start-box was opened and the lid raised, the lighting level
in the box was 5 foot candles. The ceiling of the arena was 7^2 feet from the
floor. Observation was made through two one-way mirrors. One was in a
corner by the start-box, and the other was directly opposite it. Both win-
dows were 46 in. above the floor, high enough to prevent the puppies
from seeing their reflection. Directly opposite the start-box erUrance to t c
arena, a turntable was mounted with its base to the wal . . ^

was painted red and white. A table tennis ball and a tennis a
suspended by a light chain from the top of the turntable m a
that, when revoMng, the turntable hit the balls and caused them to
bounce and the chains to emit a clanging sound. The turn able was set to
78 rpm. TTris apparatus is hereafter refened to as revo^tng stunulus^
Below the turntable was a box constructed so that 8

recorder could be placed in it from °“*«de t dynes/cm’) at the

ted a variety of noise levels ranging up to 8o dB ( • j ^ ,

start-box aLa interface. The frequendes ran^

sound was constructed so as to vary ^ “1 .. uf the

loudness in a random manner. In the ® ^ j 5 ^^d a foot above the

box entrance and to the left of 'h® ,he rag st.mulus

floor was mounted an apparatus hereafter , „gs was cotton

It consisted of a cylinder wrapped with rag . hopefully, an

rope tied so that loose ends dangled in the air m what was, pc

entidng manner for the puppy. recorded on Cramer timers

The time spent in the box ^punted on the wall to lx-

which timed to the nearest .01 mm. Ti
read outside each observation window.

272

THE dog: its domestication and behavior

water and dry food ad Ubidum. In addition, a canned food and vitamin-min-
eral mix were fed to the puppies twice daily. On days with good weather and
on which the puppy was not being tested, it was put out into a large wooded
area with other puppies its size and allowed to explore and to play for
approximately 5 hr. Water and dry food were available at all times in these
areas. Up to 30 puppies were toglher at one time. This procedure was
followed until the puppies were 10 months old.

Starting on the 21st day of age the puppies were played with and
handled in litters at least twice a week for at least 15 min each session. At
weaning each pup was assigned to an individual handler. The puppies
continued to be handled twice a week. They were walked on a leash and
exposed to many different situations and environments such as au-
tomobiles, cats, rabbits, woods, and a lake. In addition, each pup was
encouraged to “rough-house" with its handler, to play fetch, and to pull at
a rag. Once each week each puppy was subjected to a formal evaluation
session during which it underwent a battery of tests including “come" and
“sit" commands, fetch, rag play, and a maze on the 8th, 9th, and nth
weeks.

At least once each week the puppies and their handlers chased a man
(decoy) who was trailing a rag. The puppy was encouraged to bite the rag
and to pull on it. There was generally a heightened sense of excitement in
all of the puppies observing as well as the puppy doing the actual chasing.
By 12 weeks of age most puppies were vigorously pursuing the decoy and
biting the trailing rag. As the game progressed and the puppy matured,
the rag was changed to a wrapping of cloth around the decoy's arm.

To summarize the puppies' experiences prior to novel experience test-
ing, they were kennel raised, exposed to a large variety of situations, had
been taught to be somewhat aggressive with humans, had never been
scolded Of treated harshly, and had been well socialized with humans and
other puppies their size. Every effort had been made to prevent a stable
status order from developing with cage mates. If there was an obvious
dominance hierarchy developing within a pen, the pups in that pen were
separated and put into the other pens.

EQUIPMENT
Novel Situation Apparatus

The novel situation apparatus was an 8i-in. square room which had
the floor marked into nine 26 in. squares separated by 3/4-inch black tape.
The floor was painted gray and the walls were an off-white. Adjacent to
the room, hereafter called the arena, and connected directly to it by a solid

Appendix III

275

Urination and defecation while undergoing the novel experience test
was rare and subjectively appeared to be unrelated to the activity level or
distress vocalizations of the puppies. Whenever elimination did occu^
prior to the next puppy's session, the area was scrubbed and deodo^ed
with a disinfectant solution. No sanitization procedures were u ' ' e e
tween sessions of most other puppies.

Obtained blood samples were centrifuged and the day's plasma col-
lection sent to the laboratory for analysis. The EKG recordings were rea or
heart rate by counting the QRS spikes per unit of time.

TEMPERAMENT EVALUATION

Each puppy was evaluated continuously from birth. If a
nature" as hip dysplasia,

peramental character such as shyness, tear • as a breeder

appeared, the dog was eUminated ^ "og^Actual evaluaHon after
and, if the condiHon was severe, as a | who were looking

the 12th week of age was done by expenen p it may come

for a stable dog which could adapt lo^cally 0 ^ ijeals for a

across. Adapt means perform according o rurious, aggressive

given situation. Generally the ‘•“S threshold, a high pain

only when appropriate, have a h'g evaluator gave pass/fail

threshold, and show no fear in new siWa ' j|ed as to the future use
scores to all dogs at the time the deasion w ^ degrees o(

of each dog. The score was on a scaleo } o ° , .j |i,e 3 to o scores

success and 1 and o being degrees of fm • ee appropriate. If. at J

will appear throughout the results an ■ ^ en evaluation

year of age, prior disposition had not been mad^eoHhe_^p&^_^^.„,

score was also rendered on the basis o
time.

274

THE dog: its domestication and behavior

cortisol/loo ml plasma. Accuraqr calculated as the standard deviation of
percent recovery for 0.4 /xg cortisoliioo ml plasma added as internal stan-
dard was found to be 93 ± 21%. The analysis was performed by the Army
Environmental Hygiene Agency at Edgewood Arsenal, Maryland.

PROCEDURE

Not later than the day prior to testing, the hair was clipped over the left
jugular vein and the right and left fifth and sixth intercostal spaces at the
level of the elbow.

When the apparatus had been set up and tested the puppy was car-
ried into the testing area and placed on a metal table. The puppy was
positioned for bleeding from the jugular vein in such a manner as to place
minimal stress on the puppy yet aid speed and ease of collection. The
blood was collected in a heparinized 6-cc syringe with a 21-gauge iVa-in.
needle. If the collection procedure was not an immediate success with a
minimal amount of struggling by the puppy, that particular sampling at-
tempt was immediately terminated. After collection of this baseline sam-
ple, the puppy was placed in the biotelemetry harness and the electrodes
secured. The baseline heart rate was then obtained while the puppy was
minimally restrained on the table top. The telemetered EKG was recorded
for a 0.4-min period. Electrode connections and the telemetry apparatus
were observed for proper functioning while recording the heart rate. The
puppy was then immediately placed in the box and the lid secured. The
puppy was left in the box for i min. Heart rate was recorded during the
first and last 0.4-min periods while in the start-box. While the puppy was
in the start-box the arena noise, turntable, and lights were turned on.
After 1 min the guillotine door to the arena was raised and locked open.
Four different EKG recordings were made during the 3 min the puppy was
in or had access to the arena. The recordings were each 0.4 min long and
were evenly spaced throughout the 3-min session. At the termination of
the 3-min session, the puppy was called back and removed from the box.
Two minutes following termination of the lest the heart rale was again
recorded as in the baseline. At the 5-min mark after termination of the
arena testing, another heparinized blood sample was obtained from the
jugular following the same procedure used in obtaining the baseline sam-
ple. The puppy had been restrained on the table top from removal from
the box to the time of the 5-min blood collection. After that collection it
was placed alone in a large pen adjacent to the testing area. Approximately
14 min following termination of the novel experience test, the puppy was
retrieved from the pen and blood was collected at the 15-min mark.

References

Abies, E. D. (1969). Home-range studies of red foxes (Tulpis P

Mamm. 50, 108-120. . ^ ,,968) Cardiovascular changes

Adams, D. B., Bacelli, G., and Maiiaa, G. (l^h 1239-

during preparation for figftting behavio

- I r.^intnj VV. B. Saundcrs Co..
AUee, W. C. (1949). Principles of Animal tcoo^-

Philadelphia. . »he adrenal

Andrews, R. V. (1968).

metabolism of the brown lemming, y ^ behavior in nun and

Astrup, C. (1968). Pavlovian Bdiaviorm Aramab

animal. Pages 117-128 in M. W. Fox (E

W. B. Saunders. Philadelphia. ^ laboratoO'

Barnett, S. A. (1958). Experiments on "Neop o la

rats. Brit. J. Psych. 49, 195-20E „puu.)

Barnett. S. A. and Stoddart, R- ' cq 321-326.

on conflict among wild rats. /- ■ ' j^-ejopment of

Bateson, P.P.G. (1969). Imprinting and Ih £.v/y '■.-'•‘r

Pages 109-124 in A. Ambr<»c (to-t.

Academic Press, New York. 109-124- PnArrnual

Beach, F. A. and LeBoeuf, B. (1%7)- yg. 558 .

mating in the bitch. Anim. Bchau. •

References

279

Candland, D. K., Taylor, D. B., Dresdale, L., Leiphart, J. M., and Solow,
S. P. (1969). Heart rate, aggression and dominance in the domestic
chicken. /. Comp. Physiol. Psychol. S7, 70-76.

Candland, D. K., Bryan, D. C, Nazar, B. L., Kopf, K. J. and Sendor, M.
(1970). Squirrel monkey heart rate during formation of status orders, j.
Comp. Physiol. Psychol. 70, 417-423.

Cannon, W. B. (1956). "Voodoo" death. Psychosom. Med. 19, 182-190.
Chance, M. R. A. (1962). An interpretation of some agomshc
role of "cut-off" acts and postures. Symp. Zool. Soc. on . ,

Chess, S. (1969). Introduction to Child Psychiatry. Grune & Stratton, New
York.

ChiareUi, A. B. (1975). The chromosomes of

W. Fox (Ed.), The Wild Canids. Van Noshand Remhold, New YorK

• fVip pxMvations of Ihc

Qutton-Brock, J. (1969). Cami''“f , G W Dimbleby (Eds.), nr

Jericho TeU. Pages 337-^5 inP.J.UAo an.^. - ^ ^ London.

Domestication and Exploitation of Animals. VuO. i„,jnancc:a

Corbett, L. and Newsome, A. (1975). fEd ) Vie Wild Onids.

preUminary analysis. Pages 369^79. n M. W. Fox (Ed.),

Van Nostrand Reinhold, New York. ^ ^ (1975). Pel-

Corson, S. A., O'Leary Corson, ^^nderson (Ed.),Prt

facilitated psychotherapy. .

mats and Society. Balliere Tindall, on

Crisler, L. (1958). Arctic Wild. Harper Brothers, ^ ,\p-

Darwin, C. (1873). The Expression cf the Emotions m Man

pleton. New York. , Plants under Domestication.

Darwin, C. (1875). The Variation (^Animals
2nd Ed. John Murray, London. ^rnusal i^ature 211'

Dehus, J. D. (1967). Displacement acdvities and

1259-1260. . ,,Hnn input and cmooorul

Denenberg,V.H. (1964). Critical pen^^59mu^^,^^^^^, 7,,351-a--^

reactivity: A theory of infantile stim „n,o6onal reacu^ty and

Denenberg, V. H. (1967). Shmulabon mjNang.^
exploratory behavior. In . University

Neurophysiology and Emotion. Rockefe ^-ongeneuc transirax-i "

Denenberg, V. H. and Rosenberg, K- *
of information. Nature 216, 549.

27S

THE dog: rrs domestication and behavior

Beck, A. M. (1971). The life and times of Shag, a feral dog in Baltimore.

Natural Histoty 80 (8), 58-65 (October).

Beck, A. M. (1973). The Ecology of Stray Dogs: A Study of Free-ranging Urban
Animals. York Press, Baltimore. 98 pp.

Bekoff, M. (1972). An ethological study of the development of social in-
teraction in the genus Canis — A dyadic analysis. Ph.D. Dissertation,
Washington University, St. Louis, Missouri.

Belkin, D. A. (1968). Bradycardia in response to threat. Amer. Zool. 8, 775
(Abstract).

Bennett, E. L., Diamond, M. C., Rosenzweig, M. R., and Krech, D. (1964).
Chemical and anatomical plasticity of the brain. Science 146, 610-619.

Berkson, G. (1968). Development of abnormal stereotyped behaviors. De-
velop. Psychobiol 1, 118-132.

Berry, R. J. (1969). The genetical implications of domestication in animals.
Pages 207-217 in P. J. Ucko and G. W. Dimbleby (Eds.), The Domestica-
tion and Exploitation of Animals. Duckworth & Co., London.

Beylaev, D, K. and Trut, L. N, (1975). Some genetic and endocrine effects
of selection for domestication in silver foxes. Pages 416-426 in M. W. Fox
(Ed.), 77ie Wild Canids. Van Nostrand Reinhold, New York.

Blauvelt, H. (1964). The physiological analysis of aggressive behavior
(died personal communication by D. E. Davis). Pages 53-74 m W. Etkin
(Ed.), Social Behavior and Organization Among Vertebrates. University of
Chicago Press, Chicago.

Blcichcr, N. (1963). Physical and behavioral analysis of dog vocalizations.
Am. J. Vet. Res. 24, 415-427.

Bldcher, N. Personal communication. Downstalc Medical Center, Stale
University of New York, New York.

Blizard, D. A. (1971). Individual differences in autonomic responsivity in
the adult rat. Psy^toso/n. Med. 33, 445-457.

Bowlby, J. (1953). Some pathological processes set in train by early
mother-child separation. J. Mait. Sci. 99, 265-272.

Bowlby, J. (1971). Attachmait. Penguin, London.

Dmnncr, F. (1968). The application of behavior studies in small animal
practice. Pages 398-449 in M. W. Fox (Ed.), Abnormal Bdtavior in Animals.
\V. D. Saunders, Philadelphia.

Burrows, R. (1968). Wild Fax. David & Charles, Newton Abbot, England.

References

281

Fox, M. W. (1969b). The anatomy of

Canidae: a developmental and comparative study. Behmour 35,

258.

Fox, M. W. (1970a). A comparative study of !.

expressions in canids: wolf, coyote and foxes. Behmour 37, 4V

Fox, M. W. (1970b). Neurobehavioral deveiopmenUnd the geno

ronment interaction. Quart. Rev. Biol., 45,

Fox, M. W. (1971a). Integrative Development of Brain and Behmor ^n og.
University of Chicago Press, Chicago. Tnnathon

Cape, London. ,i:Hnns on the behavior of labora-

Fox, M. W. (1971c). Effects of ^“""8 , So Defining the laboratory

tory animals. Pages 294-012 m Nat. Acad. t>a.

Animal. Washington, D.C. hIp and sodo-sexual signals

Fox, M. W. (1971d). Ontogeny of sodo-mfanble and
in canids. Z. Tierpsychol. 28, 185-2iu- individual differences

Fox, M. W. (1972a). Sodo-ecologjml topUmtio^ perspective. Behmour
in wolf Utters; a developmental ana e

41, 298-313. , McCann, New York.

Fox, M. W. (1972b). Understanding Your Dog-

(Blond & Briggs, London. 1974.) Behaviour

Fox, M. W. (1973). Sodal dynamics of three ca

47, 290^01. vmir 0,1 coward. McOnn. New York.

Fox, M. W. (1974). UnderstandmYm

(Blond&Briggs, London. 1974.) ^ ^ canids. Pag“

Fox, M. W. (1975a). Evolution

in M. W. Fox (Ed.), The Wild Canids. Behavior- U

Fox, M. W. (1975b) Concepts

versity of Minnesota Press, Mmnesota. ^ ^ Anderson

Fox, M. W. (1975c). Tet-^^ B^er^Tind^aU, caiuidm

(Ed.), Pet Animals and Soaety- Toronto

Fox, M. W. (1976). The needs of people for S „et.c .

Symposium on Pets and Soaety,

Fox, M. W. (1978). fn Search of Wildness a W

tion. „ V (1973).Phy-“'‘>P“'-''

Fox, M. W. and Andrews, K. *■ '

28o

THE dog: its domestication and behavior .

Denenberg, V. H. and Whimbey, A. C. (1963). Infantile stimulation and
animal husbandry; A methodological study. }. Comp. Physiol. Psychol.
56, 877-878.

Eibl-Eibesfeldt, I. (1970). Ethology: The Biology of Behavior. Holt, Rinehart &
Winston, New York.

Eisenberg, J. and Leyhausen, P. (1972). The phylogenesis of predatory
behavior in mammals. Z. Tierpsydtol. 30, 59-93.

Ely, F. and Peterson, W. E. (1941). Factors involved in the ejection of milk.
/. Dairy Sd. 24, 211-233.

Enders, R. K. (1945). Induced changes in the breeding habits of foxes.
Sodometry 8, 53-55.

Engel, G. L. (1950). Fainting. Charles C. Thomas, Springfield, Illinois.
Epstein, H. (1971). The Origin of the Domesticated Animals of Africa. Afiicana
Publishing Corp., London.

Ewer, R, F. (1973). The Carnivores. Cornell University Press, Ithaca, New
York.

Far4 and Catlett (1971). Cardiac response and social behavior in guinea
pig. Anim. Behav. 19, 514-523.

Fox, M. W. (1964). The ontogeny of behavior and neurologic responses of
the dog. Anim. Behav. 12, 301-310.

Fox, M. W, (1965). Canine Behavior. Charles C. Thomas, Springfield, Il-
linois.

Fox, M. W. (1966). Canine Pediatrics. Charles C. Thomas, Springfield, Il-
linois. 1966.

Fox, M. W. (Ed.) (1968a), Abnormal Behavior in Animals. W. B. Saunders,
Philadelphia.

Fox, M. W. (1968b). Use of the dog in behavioral research. Pages 27-80 in
W. 1. Gay (Ed.), Methods of Animal Experimentation, Vol III. Academic
Press, New York.

Fox, M. W. (1968c) The influence of domestication on the behavior of
animals. In M. W. Fox (Ed.), Abnormal Behavior in Animals. W. B. Saun-
ders, Philadelphia.

Fox, M. W. (1968d). Socialization, environmental factors and abnormal
behavorial development in animals. Pages 332-355 in M. W. Fox (Ed.),
Abnormal Behavior in Animals. W. B. Saunders, Philadelphia.

Fox’, M. W. (1969a). Ontogeny of prey-killing behavior in Canidae. Behavior
35, 259-272.

References

283

Gray, A. P. (1954). Mammalian hybrids. A check-list with bibliography.
Technical Communication 10, Comm. Agriculture Bureaux, F am
Royal, Bucks, England.

Guthrie, R. D. (1975). A hypothesis of density-adapted mo^hs among
Northern canids. Pages 414^15 In M. W. Fox (Ed.), The Wild Camds.
Van Nostrand Reinhold, New York.

Hale, E. B. (1962). Domestication and the ^ Co^'

Hafez (Ed.), The Behavior Domestic Animals. Williams

Baltimore. p

Hale, E. B. (1969). Domestication and the M

22^ ill E. S. E. Hafez (Ed.), The Behavior of Domestic Anima

edition. Balliere Tindall, London. n 68 74

Harlow, H. F. (1959). Love in infant monkeys. Sd. Am. 20 ,

Hediger. (1950). Wild Animals in Captivity.

Hediger, H. (1955). Studies of the Physiology and Behavwu

in Zoos and Circuses. Butterworth, London.

Henderson, N. (1970). Cerie 6 c 72 , 505-5^

obscured by laboratory reanng. J. Co p- y TPoroduction.

Hill, J. L. (1974). Peromyscus: Effect of early pairmg

Science, 186, 1042-1044. during sudden pro-

Hofer, M. A. (1970). Cardiac and 32, 633-647.

longed immobility in wild rodents, sy primate research

Jensen, G. D. and Bobbitt, R. A. (1968)- J"’P''JX'serman (Ed.), Sririirr
for understanding infant New York.

and Psychoanalysis, Vol. 12. Grune ' „ Press, New

Ioffe, ,. M. (1969h Prenatal Determinants cf Behavior. Pergam

Tork. timber wolf (O'”'*

JosUn, P. W. B. (1966). Summer A'*”“'^°>[l^°thesis, University of o-
packs in Algonquin Park. Unpublished M.s.

ronto. 99 pp. d the legacy of ton '

Keeler, C. (1970). Melanin, adrenalin, an

81-88. , . -n color phases of the red

Keeler, C. (1975). Genetics of behavior Q^mds. Van Nost

fox. PagL 399-413 in M. W. Fox (Ed.),

Reinhold, New York. coyote-dog h) n

Kennelly, J. J. and Roberts, J. D* t
Mammal. 50, 830-831.

282

THE dog: its domestication and behavior

correlates of individual differences in wolf litters. Behaviour XLVI, 129-
140.

Fox, M. W. and Bekoff, M. (1975). The behaviour of dogs. Pages 370^09
in E. S. E. Hafez (Ed.), The Behaviour of Domesticaled Animals, 3rd edition.
Balliere Tindall, London.

Fox, M. W. and Clark, A. L. (1971). Ontogeny and temporal sequencing of
agoiusdc behavior in the coyote, {Canis latrans). Z. Tierpsychol. 28, 262-
278.

Fox, M. W. and Cohen, J. A. (1977). Canid communication. In T. A. Sebeok
(Ed.), How Afiimals Communicate. Indiana University Press, Blooming-
ton, Indiana. In press.

Fox, M. W. and Spencer, J. (1%9). Development of exploratory behavior in
the dog: Age or experience dependent? Develop. Psychobiol. 2, 68-74.

Fox, M. W. and Walls, S- (1973). Wild animals in captivity: Veterinarian's
role and responsibility . }. Zoo Anim. Med. 4(3), 7-17.

Fox, M. W., Folk, G. E., and Folk, M. (1970). Physiological differences
between alpha and subordinate wolves in a captive sibling pack. Amer.
Zool. 10, 487.

Fox, M. W., Lockwood, R., and Shideler, R. (1974). Introduction studies in
captive wolf packs. Z. Tierpsydtol. 35, 39-48.

Funkenstein, D. H. (1955). The physiology of fear and anger. Sd, Am.

Gantt, W. H. (1944). Experimental Basis for Neurotic Behavior. Hoeber, New
York,

Gantt, W. H., Newton, J. E. O., Royer, F. L., and Stephens, J. H. (1966).
Effect of person. Cond. Reflex 1, 18-35.

Gellhom, E. (1968). Central nervous system tuning and its implications for
neuropsychiatry. /. Neru. & Ment. Dis. 147, 148-162.

Gier, H. T. (1968). Coyotes in Kansas. Kansas Agricultural Experiment Sta-
tion Bulletin 393, 1-97,

Gier, H. T. (1975). Ecology and social behavior of the coyote. Pages 247-
262 in M. W. Fox (Ed.), The Wild Canids. Van Nostrand Reinhold, New
York.

Ginsburg, B, E. (1968). Genotypic factors in the ontogeny of behavior. In J,
H. Masserman (Ed.), Animal and Human. Grune & Stratton, New York.

Graham, F. K. and Clifton, R. K. (1966). Heart-rate change as a component
of the orienting response. Psychol. Bull. 65, 305-320.

References

283

Lor^, K ( 1 ^ 0 ). Studies in Animal and Human Behavior, Vol. II. Harvard
University Press, Cambridge, Massachusetts.

™ Van

nJostrand Remhold, New York.

Lucas E. A., Powell, E. W., and Murphree, O. D. (1974). Hippocampal
tneta m nervous pointer dogs. Physiol. & Behav. 12, 609-613.

Lynch, J. J, (1970). Psychophysiology and development of sodai attach-
ment. /. Nerv. & Ment. Dis. 151, 231-244.

Lynch, j. J, and Gantt, W. H. (1968). The heart rate component of the
sodal reflex in dogs: the conditionai effect of petting and person. Cond.
Reflex 3, 69-80.

Macintosh, N. W. G. (1975). The origin of the dingo; An enigma. Pages
87-106 in M. W. Fox (Ed.), The Wild Canids. Van Nostrand, ReinhoJd,
New York.

Marler, P, and Hamilton, W. J-, HI. (1966). Mechanisms of Animal Behavior,
John Wiley & Sons, New York.

Mason, W. A. (1967). Motivational aspects of sodal responsiveness in
young chimpanzees. Pages 12-^ in H. W. Stevenson, E. H. Hess
and H. L, Rheingold (Eds.), Early Behavior — Comparative and Development
tal Approaches. John Wiley & Sons, New York.

McBride, R, L,, Klemm, W. R., and McGraw, C P. (1969). Mechanisms of
the immobility reflex ("animal hypnosis"). Comm. Behav. Biot. 3, 33-59.

McNab, B. K. (1963). Bioenergetics and the determination of home range
size. Amer. Natur. 97, 133-140.

Mech, L. D. (1970). The Wolf. Natural Histoiy Press, Neiv York.

Meggitt, M. J. (1965). The association between Australian aborigines and
dingoes. Pages 7-26 in A. Leeds and A. P. Vayada (Eds.), Man, Culture &
Animals. Publ. #78, AAAS, Washington D.C
Meier, G. W. (1961). Infantile handling and development in Siamese kit-
tens. /. Comp. Physiol. Psychol. 54, 284-286.

Mengel, R. M. (1971). A study of dog-coyote hybrids and implications
concerning hybridization in Canis. /. of Mamm. 52, 316-336.

Menzel, R. and Menzel, R. (1948). Observations on the pariah dog. Pages
968-990m B. Vesey'PitzQerald (Ed.),The Bookof the Dog. Borden Publish-
ing, Toronto, Canada.

Meyer-Holzapfel, M. (1968). Abnormal behavior in zoo animats Paf;es

284

THE dog: its domestication and behavior

King, ]. A. (1968). Spedes-spediidty and early experience. Pages 42-64 m
G. Newton and S. Levine (Eds.), Early Experience and Behavior. Qiarles
C. Thomas, Springfield, Illinois.

Klinghammer, E. (1967). Factors influencing choice of mate in altridal
birds. Pages 297-303 in H. W. Stevenson, E. H. Hess, and H. L. Rhein-
gold (Eds.), Early Behavior. Academic Press, New York.

Kolenosky, G. B. (1971). Hybridization between a wolf and a coyote. ].
Matnmal. 52, 446-449.

Kovach, J. K. and Kling, A. (1967). Mechanisms of neonate sucking be-
havior in the kitten. Anim. Behau. 15, 91-101.

Kummer, H. (1971) Primate Societies. Aldine, Chicago.

Kuo, Z. Y. (1960). Studies on the basic factors in animal fighting. VII.

Interspedes coexistence in mammals. /. Genet. Psychol. 97, 211-225.
Kurtsin, I. T. (1968). Pavlov's concepts of experimental neurosis and ab-
normal behavior in animals. Pages 77-106 in M. W. Fox (Ed.), Abnormal
Behavior in Animals. W. B. Saunders, Philadelphia.

Lacey, J. I. and Lacey, B. C. (1970). Some autonomic-central nervous sys-
tem interrelationships. Pages 205-226 in P. Black (Ed.), Physiological Cor-
relates of Emotion. Academic Press, New York.

Levine, S. and Mullins R. F. (1966). Hormonal influences on brain organi-
zation in infant rats. Science 152, 1585-1592.

Leyhausen, P. Verhaltensstudien an Katzen. PaulParey, Berlin and Hamburg,
1973.

Liddell, H. S. (1954). Conditioning and emotions. Scientific American 190,
48-57.

Linn, J. (1974). An evaluation of puppy heart rate and plasma cortisol
levels as temperament predictors in German shepherd dogs. Unpub-
lished M.Sc. thesis, Washington University, St. Louis, Missouri.
Lockwood, R. (1976). An ethological analysis of sodal structure and affilia-
tion in captive wolves. Unpublished Ph.D. thesis, Washington Univer-
sity, St. Louis, Missouri.

Long, E. M., Truex, R. C., Friedmann, K. R., Olson, A. K., and Phillips, S.
J. (1958). Heart rate of the dog following autonomic denervation. Anat.
Rec. 130, 73-89.

Lorenz, K. (1966). On Aggression. Harcourt, Brace & World, New York.

Lorenz, K. (1968). Evolution and Modification of Behavior. University of
Chicago Press, Chicago.

References

287

monkeys In M. W. Fox (Ed.), Abnormal Behavior in Animals. W. B. Saun-
aers, Philadelphia,

Sackett, G. P., Porter, M., and Holmes, H. (1965). Choice behavior in
rnesus monkeys effect of stimulation during the first month of life
Sciefice 147, 305-306.

Salzen, E. A. and Comell, J. M. (1968). Self-perception and species recog-
nition in birds. Behaviour 30, 44-65.

Schenkel, R. {1947). Expression studies of wolves. Behaviour 1, 81-129.

Schneirla, T. C. (1959). An evolutionary and developmental theory of
biphasic processes underlying approach and withdrawal. Pages 1-42 in
M. R. Jones (Ed.), Nebraska Symposium on Motivation. Nebraska Univer-
sity Press, Lincoln, Nebraska.

Schneirla, T. C. (1965). Aspects of stimulation and organization in
approach/withdrawal processes underlying vertebrate behavioral de-
velopment. Pages 1-74. in D. S. Lehrman, R. A. Hinde, and E. 5haw
(Eds.), Advances in the Study of Animal Behavior, VoJ. I. Academic Press,
New York.

Scholander, P. F., Irving, L., and Grinnell, $. W. (1942) Aerobic and
anaerobic changes in seal muscles during diving. /. Biol. Chem. 142,
431-440.

Schutz, F. (1965). Sexuelle Pragung bei Anatiden, Z. Tierpsychol. 22, 50-
103.

Scott, J. P. (1962). Critical periods in behavioral development. Science
138, 949-958.

Scott, J. P. {1968a). Evolution and domestication of the dog. Pages 243-275
in T. Dobzhansky, M. K. Hecht and W. C. Steere (Eds.), Evolutionary
Biology, Vol. II. Academic Press, New York.

Scott, J. P. (1968b) Early Experience and the Organization of Behavior. Brooks'
Cole, Belmont, California.

Scott, J. P. (1973). Animal Behavior. Chicago: Chicago University Press.
1973.

Scott, J. P. and FuUer. J. L. (1965). Genetics and Social Behmiar of the Dos
University of Chicago Press, Chicago.

Scott, M. D. and Causey, K. (1973). Ecology ol feral dogs in Alabama /
WUdl. Mgmt. 37, 252-265.

Silber, K. H.,
estimation

Bush R. D., and Oslapas, H. (I95S). Practical procedure for
of corticosterone and hvdiocorh'sone. Clm. Chew. 4, 27a-2&>

i86

THE dog: its domestication and behavior

476-503 in M. W. Fox (Ed-). Abnormal Behavior in Animals. W. B. Saun-
ders, Philadelphia.

Morton, J. R. (1968). Effects of early experience on behavior. Pages 261-262
in M. W. Fox (Ed.), Abnormal Behavior in Animals. W. B. Saunders,
Philadelphia.

Murphree, O. D., Dykman, R. A., and Peters, ], E. (1967). Genetically-
determined abnormal behavior in dogs. Cond. Reflex 4, 199-205.

Murphree, O. D., Peters, J. E., and Dykman, R. A. (1969). Behavioral
comparisons of nervous, stable, and crossbred pointers at ages 2, 3, 6, 9
and 12 months. Cond. Reflex 4, 20-23.

Nesbitt, W. H. (1975). Ecology of a feral dog pack on a wildlife refuge, in
M. W. Fox (Ed.), The V/ild Canids. Van Nostrand Reinhold, New York.

Newton, ]. E. O. and Gantt, W. H. (1968). The history of a catatonic dog.
Cond. Reflex 3, 45-61.

Newton, J. E. O., Murphree, O. D., and Dykman, R. A. (1970). Sporadic
transient atrioventricular block and slow heart rate in nervous pointer
dogs. Cond. Reflex 5, 75-89.

Nottebohm, F. (1970). Ontogeny of bird song. Science 167, 950-956.

Obrist, P. A. (1968), Heart-rate and somatic motor coupling during classi-
cal aversive conditioning in humans. J. Exp. Psychol. 11, 180-193.

Pavlov, I. P. (1928). (Gantt, W. H., transl.) Lectures on Conditioned Reflexes.
International Publishers, New York.

Richter, C. P. (1954), The effects of domestication and selection on the
behavior of the Norway rat, /. Nat. Cancer Inst. 15, 727-738.

Richter, C. P. (1957). On the phenomenon of sudden death in arumals and
man. Psychosom. Med. 19, 191-198.

Riesen, A. H. (1961). Stimulation as a lequiremenl for growth and function
in behavioral development. Pages 57-80 in D. VV. Fiske and S. R. Maddi
(Eds.), Functions of Varied Experience. Dorsey Press, Homewood, Illinois.

Rouget, Y. (1970). Personal communication. Department of Ethology,
University of Rennes; film presented at International Ethology Confer-
ence.

Royer, F. L. and Gantt, W. H. (1961). The effect of different persons on the
heart rale of dogs. Paper presented at Eastern Psychological Association,
Philadelphia, March 1961.

Sackett, G. P, (1968). Abnormal behavior in laboratory-reared rhesus

References

289

monkey calls under normal conditions and under acoustic isolation.
Behaviour XLVJI, Part 3^, 230-239,

Wolf, S. (1964). The bradycardia of the dive reflex — a possible mechanism
of sudden death. Trans. Amer. Clin. Climat. Ass., 192-200.

Woolpy, J. H. Socialization of wolves. Pages 82-94 in J. H- Masserman
(Ed.), Animal & Human. Grune & Stratton, New York.

Woolpy, J, H. and Ginsburg, B. E, (1967), Wolf sodalization: a study of
temperament in a wild social s^pedes. Amer. Zool. 1, 357-364.

Zeuner, F, E. A History of Domesticated Animals. (1963), Harper iSc Row,
New York.

Zimen, E. (1971). Wol/e und Konigspudel, In W, Wickler (Ed,), Ethol.
Studien. Piper Verlag, Munchen,

288

THE dog: its domestication and behavior

Silver, H. and Silver, W, T. (1969). Growth and behavior of the coyote-like
canid of Northern New England with observations on canid hybrids.
Wildlife Monographs 17, 1-41.

Spitz, R. (1949). The role of ecological factors in emotional development.
Child Development 20, 145-155.

Spitz, R. (1950) . Anxiety in infancy: a study of its manifestations in the first
year of life. Int. J. Psychoatial. 31, 138-143.

Staines, H. J. (1975). Distribution and taxonomy of the Canidae. Pages 3-26
in M. W. Fox (Ed.), The Wild Canids. Van Nostrand Reinhold, New York.

Tembrock, G. (1958). Lautenwiclung bein fuchs; sichter gemacht.
Umschau, 58, 566.

Tembrock, G. (1960). Spezifische laulformen beim rotfuchs (vulpes vul-
pes) und ihre beziehunger zum verhalten. Saugertierkundl. Mitt 8, 150-
154.

Tembrock., G. (1968). Land mammals. Pages 338-404 in T. A. Sebeok
(Ed.), Animal Communication. Indiana Unversity Press, Bloomington, In-
diana.

Theberge, J. B. and Falls, J. B. (1967). Howling as a means of communica-
tion in wolves, Amer. Zool. 7, 331-338.

Thomas, A., Chess, S., and Birch, H, G. (1970). The origin of personality.
Sci. Amer. 223, 102-109.

Thompson, W. R. (1957). Influence of prenatal maternal anxiety on emo-
tional reactivity in young rats. Science 125, 698-699.

Trumler, E. (1973) Your Dog and You. Seabury, New York. 1973.
Vandenberg, J. G. (1969). Male odor accelerates female sexual maturation
in mice. Endocrinology 84, 658-660.

Van Lawick, H. and Van Lawick Goodall, J. (1971). ITie Innocent Killers.
Houghton Mifflin, New York.

Vauk, G. (1953) Die abwandlung der Beutefanghandlung des hundes im
zuge der domestikation. Zool. Anzeiger (Deutsche Zoologische Geselscaft),
Supplement 17, 180-184.

Volgyesi, F. A. (1966). Hypnosis of Man and AnUnal. Ballifere, London.

Wallace, R. K. and Benson, H, (1972) The physiology of meditation. Sci.
Amer. 226, 84-90.

Weber, E. and Weber, E. H. (1845). Gled in Long et al. (1958).

W'mter, P., Handley, P. Ploog, D. and Schott. (1973). Ontogeny of squirrel

Index

Abies, E. D., 51
Abnormal behavior, 237, 259
Activity, feral dogs, 48
Adams, D. B., 178
Adaptive radiation, 7
Adrenals and stress, 202
A-elidting sounds, 84
Affection and physiology/ 231
Aggression, development of, 33 ^
displays, 98, 99; and prey killing,
130; and sex, 147, 149
AUee, W. C., 4
Ancestry, dog, 242
Andrews, R. V., 269
Astrup, C., 226
Attachment. See socialization
Attention bradycardia, 179
Autonomic balance, 226
Autonomic tuning, 165

Barks, 76, 87
Bamclt, S. A., 15, j6, 18
Bateson, P. P. G., 172
Beach, F. A.. 37

-k, A. M., 65

savior, development, .55: ^
,me 5 tication, 8 , 25 o;ofh)‘’"'^*'

coff, M., 9 >

kin, D. A., ^ 77 ; -“5

melt. E. L., 166, 169
■kson, G.. 188

TV, R- ^9

6, 127

uvelt, H-, 133

ichcr, N-. 7-’ "5

^rd, D A . 190

,d, scxuJl, 28

Wby. J . '?;* ...4

dvcjrdu '

fn

in 2nd vnvnvnn-.n .
cd diift-vcnco. 3 ,

..duTg J j

•ding. ‘ jr-!

nHn>L

Index

293

Environment, deprivation, 166;
and development, 171
Environmental influences, 153
Environmental input, 175
Epstein, H., 242, 243
Ergotropic tuning, 205
Estrus, 37; hybrids, 102
Ethogram, 239
Ethostasis, 232

Evolution of submission, 229
Ewer, R. F., 130

Fara, G., 178
Fear bradycardia, 177
Fear of strangers, 38
Feral dogs, 41; activity of, 41
Flight distance, 11, 216
Food and sociability, 243
Forelimb stab action, 123
Fox, M. W., 8, 14, 16, 24, 28, 32,
36, 41, 66, 69, 70, 77, 85, 87, 95/
97/ 99/ 102, 104, 105/ ^^ 7 '

13O/ 135/ 154/ 158/ 161, 164, 167,
171, 198, 200, 203, 218, 23I/ 235,
238, 250, 259, 261
Funkenstein, D. H,, 207

Gantt, W. H., 182
Gellhom, E., 178, 205, 208, 222,
226

Genetics, of behavior, 91; and
domestication, 238; and envi-
ronment, 169; and handling, 164;
and heart rate, 227
"Gentling," 162
Gier, H. T., 28, 104
Ginsburg, B. E., 164, 171, 195
Graham, F. K., 181, 228
Gray, A. P., 36
Grooming, effects, 162
Grunts, 77
Guthrie, R. D., 237

Hale, E. B., 3, 5, 14, 15/ 16, 17/ 18
Handler, reactions to, 142
Handling, and domestication, 165;

effects, 163; in pups, 192
Handling responsiveness, 102
Handling stress, 189; in dogs, 164
Harlow, H. F., 184
Heart rate, development, 184; and
early experience, 189; and tem-
perament, 34, 187, 196, 209
"Heck's law," 237
Hediger, H., 150, 157, 216
Henderson, N., 169
HiU, ]. L., 170
Hofer, M. A., 177/ 228
Home range movements, 51
Howls, 77

Hunting behavior, 61, 107
Hunting, breed differences, 129
Hybridization, 17
Hybrids, and prey kiUing, 107;
specific actions, 123

Imprinting, 158
Inbreeding problems, 257

fandlism, 151
euinal response, 102
heritance, non-genehc, 171

;n, G. D., 168
, J. M., 186, 19°

r, C., 16, 237
elly, J- J -
g of prey, 107
J. A., 17=

hammer, E , 150, 1S7

Kovach, J. K , 17*
Kummer, H , 3b

292

THE dog: its domestication and behavior

Brunner, F., 259
Burrows, R., 28, 123

Candland, D. K., 34, 204, 206
Canids, distribution, 22; social or-
ganization, 26, 28; taxonomy, 23
Cannon, W. B., 225
Cardiac activity, 177
Chance, M. R. A., 224, 231
Chess, S., 161
ChiarelU, A. B., 242
Clutton-Brock, J., 249
Coat color and temperament, 16,
237

Coat type, hybrids, 95
Communication, in canids, 24;
vocal, 69

CompatabUity, interspecies, 137
Conditioning and heart rate, 182
Contactual comfort, 183
Coos, 77

Corbett, L., 39, 150

Corson, S. A., 259

Coyote X dog hybrid, 91

Crisler, L., 74

Critical distance, 216

Critical period, socialization, 160

Critical periods, 171

Darwin, C., 85
Degeneration, 240, 258
Delius, J. D., 229
Denenberg, V. H., 162, 163, 171,
172, 179, 189, 190, 195
Dependence, 253
Deprivation, environmental, 166
Destabilization, 19
Development, heart rate, 184;
stages of, 153; vocal, 80
Deviant behavior, 17
Dingo, 241, 249

Disease susceptibility, 225
Disinhibition of bite, 129
Displays, and domestication, 236;
and heart rate, 230
Diving reflex, 178
Docility, 8

Dog X coyote hybrids, 91
Dog, domestication, 37; origins,

36

Dogs, feral, 41

Domestication, behavior effects,

8; comparisons, 37; environmen-
tal influences, 153; history of, 5;
and hunting, 129; and in-
terspecies interactions, 138;
physiological effects, 18; predis-
posing factors, 9; and prey catch-
ing, 107; stages of, 5; and sympa-
thetic tone, 222; and vocaliza-
tions, 69

Dominance and heart rate, 198,
204

Early experience, 162; and play,
138

Early rearing patterns, 158
Ecology, and behavior, 26; feral
dogs, 41

Economic charactercstics, 8
Ectomorph, 93, 95
EEG development, 155
Eibl-Eibesfeldt, I., 14
Eisenberg, J., 109
Ely, F., 15

Empathy and heart rate, 232
Emotional development, 175
Emotional disorders, 161, 259
Emotions and heart rate, 177
Enders, R, K., 11
Endomorph, 95
Engel, G. L., 178
Enrichment, environmental, 166

Index

Environment, deprivation, 166;
and development, 171
Environmental influences, 153
Environmental input, 175
Epstein, H., 242, 243
Ergotropic tuning, 205
Estrus, 37; hybrids, 102
Ethogram, 239
Ethostasis, 232
Evolution of submission, 229
Ewer, R. F., 130

Fara, G., 378
Fear bradycardia, 177
Fear of strangers, 38
Feral dogs, 41; activity of, 41
Flight distance, 11, 2x6
Food and sociability, 243
Forelimb stab action, 123
Fox, M. W., 8, 14, 16, 24, 28, 32,
36, 41, 66, 69, 70, 77, 85, 87, 95,
97/ 99/ ^02, 104, 105, 107, 109, 126,
130, 135, 154, 158, 361, 164, 167,
171, 198, 200, 203, 2x8, 231, 235,
238, 250, 259, 261
Funkenstein, D. H., 207

Gantt, W. H., 182
Gellhorn, E., 178, 205, 208, 222,
226

Genetics, of behavior, 91; and
domestication, 238; and envi-
ronment, 169; and handling, 164;
and heart rate, 227
"Gentling," 162
Gier, H. T., 28, 104
Ginsburg, B. E., 164, 171, 395
Graham, F. K., 181, 228
Gray, A. P., 36
Grooming, effects, 162
Grunts, 77
Guthrie, R. D., 237

293

Hale, E. B., 3, 5, 14, 15, 16, 17, 18
Handler, reactions to, 142
Handling, and domestication, 165;

effects, 163; in pups, 192
Handling responsiveness, X02
Handling stress, 189; in dogs, 164
Harlow, H. F., 184
Heart rate, development, 184; and
early experience, 189; and tem-
perament, 34, 187, 196, 209
"Heck's law," 237
Hediger, H., 150, 157, 216
Henderson, N., 169
Hill, J. L., 170
Hofer, M. A., 177, 228
Home range movements, 51
Howls, 77

Hunting behavior, 61, 107
Hunting, breed differences, 129
Hybridization, 17
Hybrids, and prey killing, 107;
specUic actions, 123

Imprinting, 156
Inbreeding problems, 257
Infantilism, 353
Inguinal response, 102
Inheritance, non-genetic, 171
Interspedes play, 133

Jensen, G. D., 168
Joffe, J. M., 165, 166, 190
Joslin, P. W. B., 72

Keeler, C., 16, 237
Kennelly, J. ]•> 3^4
KiUing of prey, 107
King, }. A-, 172

KHnghammer, E.. 150, i 57
Kolenosky, G. B.. 36
Kovach, /• K ' ^72
Kummer, H.. 36

294

the dog: its domestication and behavior

Kurtsin, I. T., 179, 205, 222, 225,
226

Lacey, J. L, 179, 180, 181, 189, 204,
205, 218, 228, 229
Leadership, 52
Leap-stab action, 123
Levine, S., 163, 174, 189
Leyhausen, P., 123, 224, 226
Licking, social, 102
Liddell, H. S., 183
Linn, J., 208, 209
Lockwood, R., 237
Long, E. M., 177
Lorenz, K., 16, 157, 251
Lucas, E. A., 228
Lynch, J. J., 182, 183, 184, 204, 205

Macintosh, N. W. G., 24, 249
Man-dog relations, 257
Marler, P., 25
Mason, W. A., 168
Marking behavior, 52
McBride, R. L., 221
NcNab, B. K., 51
Mech, L. D., 29
Meggitt, M. J., 241
Meier, G. W., 164, 194
Mengel, R. M., 91, 104
Menzel, R., 245, 247, 251
Mesomorph, 95
Metacommumcation, 66
Mews, 77

Meyer-Holzapfel, M., 168, 172
Mimicry, 236
Mixed sounds, 77
Monogamy, 37
Morning greeting, 60
Morphology, canid hybrids, 92
Morton, J. R., 174
Mother and stress, 183
Mothering, effects, 163

Murphree, O. D., 204, 227
Muzzle-bite, 250

Needs, human, 257
Neophobia, 237
Neoteny, 9, 151

Nervous system, development,
175

Nervous topologies, 222, 225
Nervous typology, 205
Nesbitt, W, H., 41, 67
Newton, J. E. O., 228
Noltebohm, F., 174

Ontogeny. See development
Orbrist, P. A., 228
Organization of behavior, 238
Origms of dog, 241
Overattachment, 161
"Overbreeding," 257
Overindulgence, 161, 260

Pack cohesion, 61
Paedomorphic features, 8
Parasites, 45

Parasympathetic stimulation, 224
Parental behavior, 11
Pariah dogs, 244
Pavlov, I. 181

"Perpetual puppy" syndrome, 161
Petting, and heart rate, 182; and
stress, 183
Phenodeviants, 238
Phenotype, changes, 171; domes-
tic fitness, 169; enhancement, 195
Phylogeny of submission, 229
Plasma cortisol, 202, 209
Play, in canids, 133
Predation, 27
Prenatal influences, 165
Prey-catching, 107
Prey, dissection of, 116, 119

Index

295

Primary social attachments, 145
Psychophysiological dependency,
231

Psychosomatic disorders, 258,
259, 260

Reactivation, 238
Relationships, man-dog, 257
Reproduction and socioenviron-
ment, 170

Richter, C. P., 14, 18, 177, 223, 226
Royer, F. L., 182

Sackett, G. P., 158, 167
Salzen, E. A., 158
Schenkel, R., 80

Schneirla, T. C., 83, 162, 205, 224,

231

Scholander, P. F., 178
Schutz, F., 157

Scott, J. P., 37, 41, 143' ^56' ^59'
171, 172, 174, 179/ 182, 188, 189,
236, 243, 247, 248/ 250
Scott, M, D., 51
Screams, 76

Secondary socialization, i43'
Selective breeding, 15
Sensitive periods, 171
Sex, and environment, 170; and
sociability, 147
Sexual behavior, 11, 37
Sexual maturity, 148, 253
Shyness, 95, 97
Silber, R. H., 269
Silver, H., 91

Sociability and territory, 246
Social behavior, feral dogs, 5®
Social bonds. See socialization
Social deprivation, 159
Social homeostasis, 232
Social inhibition of aggression, 13°
Social organization, 24, 26

Social regression, 159
Socialization, 14, 39, 141, 156/ 160;
of dog and cat, 158; and heart
rate, 212; and touch, 184
Socioecology, 21
Sociosexuai discriminations, 149
Somatotypes, hybrids, 95
Sounds, vocal, 72
Speciation, artificial, 7
Species-typical actions, 139
Spermatogenesis, seasonal, 253
Spitz, R., 162, 184, 224
Staines, H., 21
Stereotypes, 168
Strangers, fear of, 142
Stress, handling, 189
Stress resistance, 164, 189
Structure and domestication, 248
Submission, evolution of, 229;
and heart rate, 219
Sudden death, 225
Supracaudal gland, 95
Sympathetic tone, 34

il gland, 251
il posture, 250
ming, 3' 4'

<onomy of canids, 23
aibrock, G., 7^' 80, 5 '
„perament, and 3 "

,d handling, 165; “nd hurt

ite, 177, 209,

jterogeneity. 32
nperature and actrvny, 48

nporal sequencing m prey Wl

IK, 107, 120

nporary packs, 84 - hs

Titoriality, 246

;bcrge, J- B ■ 7 -

,ntas, A , 165. ' 97 ,

rmpson, W R , .hi

■eshold of acuoni, .-h

THE dog: rrs domestication and behavior

296

Timidity, 95, 97; and heart rale,
204

Tonic immobility, 183, 224, 229
Tooth eruption, hybrids, 95
Trophotropic tuning, 205
Trumler, E., 237

Urination, social, 52

Vagal inhibition, 185
Vandenberg, J. G., 170
Van Lawick, H., 28, 35
Vauk, G., 128

Vocalizations, 25, 69; group, 29
Vdlgyesi, F. A., 229

Wallace, R. K., 232
Weber, E., 177
W-eliciting sounds, 84
Whines, 74

Wild canids, behavior, 21; taming
of, 141

Wildness, 95
Winter, P., 86
Wolf, S., 178
Wolf ancestry, 248
Wolf-dog comparisons, 253
Woolpy, J. H., 14, 143, 159

Yelps, 75

2 euner, F. E., 5, 6
Zimen, E., 151